PDA

View Full Version : Origins of M222, DF23, the NW Irish and connections to Niall of the Nine Hostages



TigerMW
01-21-2014, 03:10 PM
I don't know where DF49 and DF23 originated and I don't know if M222 is really connected to Niall, ancient Irish King. Its a thought provoking conversation, though.

Here is some new information. One does not a trend make, but here is an interesting new DF23+ test result:

286325 Martin Heil, b.1877, Baranya, Southern Transdanubia, Hungary

I've seen him in the L21 project for some time but was always perplexed. I didn't see any SNP testing until now. My spreadsheets would pick him as a potential DF49* or DF23* type, but every time I would check GD's they'd be so far out that I'd discard his haplotype from the spreadsheet updates.

He's got Geno 2 results now, and sure enough, he is DF23+. I don't see an M222+ so I presume he is M222- but we should look at his raw results to be sure.
http://www.familytreedna.com/public/R-L21/default.aspx?section=ysnp

Hungary is a quite a ways from Ireland but we do have some cases of people traveling quite widely in our historic period. However, I think the other factor is the STR signature. He has a high DYS481 value and DYS413 mutations which are typical in M222 and some parts of DF23. However, the point is that the rest of his STRs scarely match any clusters.

His closest GDs to anyone within L21 are GD=15 @67 which means he broke away from the rest of DF23 a long time ago, perhaps about the time of Caesar.

His closest GD's (15) are:

f286325 Heil R1b-L21>DF13>DF49>S476>DF23 DF23+ L21+ L21+ Z290+ (G2) 49-23*- uas
f96218 Grenon zzL21suspect 49- uas (France)
f95659 Johnston zzL21suspect 49-2329-1922-J
f101298 MacGregor zzL21suspect z43811-10
f36459 McKenzie zzL21suspect 49-23*-1922-C

Grenon is just suspected L21 but he does match some of those signature markers for DF23 types, which is why I included him in the file.

Of course, the Heil surname could relate to German immigration east into Hungary.

rms2
01-21-2014, 03:36 PM
DF23, Z2961, and even M222 are known on the Continent, especially France, but, of course, the tendency has always been to dismiss them as the descendants of "Wild Geese" or of Irish monks for whom celibacy was merely a part-time commitment.

Over on that thread about Jean's book, you mentioned her idea that M222 could represent the arrival in the Isles of some continental settlers bearing La Tene culture and skills. Maybe that's right, or it began with DF23.

TigerMW
01-21-2014, 03:54 PM
...
Over on that thread about Jean's book, you mentioned her idea that M222 could represent the arrival in the Isles of some continental settlers bearing La Tène culture and skills. Maybe that's right, or it began with DF23.

Here it is. In "Ancestral Journeys...", Jean Manco wrote,
"A more likely genetic signature of La Tene in Ireland is Y-DNA R1b-M222, carried by up to 44 per cent of men in parts of Northern Ireland today.
...
This is not the pattern we would expect from Irish migrants into Britain. So R1b-M222 hints at La Tène movements into Ireland."

I'd have to go back and read through all of that again, but I think the general idea is that M222 came into Ireland from Northern Britain with La Tène. Ironically, that would mean it came in as some kind of P Celtic group that integrated into a Q Celtic based Gaelic (to be) society.

Wasn't Niall supposed to come from the Connachta, which would be westerly in Ireland? The idea of a La Tène immigration seems to be at odds with that. Perhaps I have my Irish stories mixed up.

Dubhthach
01-21-2014, 03:59 PM
He should be sponsored for Z2961. After all we know that this is the specific branch of DF23 that M222 sits in, an upgrade to 111 markers would be good as well. I'll just cross-post what I wrote on the L21 mailing list:



---------- Forwarded message ----------
From: Paul Ó Duḃṫaiġ <[email protected]>
Date: Tue, Jan 21, 2014 at 2:57 PM
Subject: Re: [R1b-L21-Project] Heil from Hunary is DF23
To: [email protected]


Heil might be a german surname. There was a fairly large ethnic German population in Hungary and Central Europe in general (Danube Swabians for example). Of course after World War II about 12 million ethnic German's were expelled from Central/Eastern Europe.

I have a feeling this is the same Heil -- 1877 in the post --
http://archiver.rootsweb.ancestry.com/th/read/DONAUSCHWABEN-VILLAGES/2013-01/1359216121

-Paul
(DF41+)



That rootsweb message talks about a family folklore of origin in Alsace-Lorraine. So basically the wider Rhineland, there's been some talk before about the general Rhineland/Eastern France as been potential source location for the likes of M222 in the past (or probably more likely DF49/DF23 etc.).

Niall is a "quasi-historical" figure. In general historians do agree that he existed but alot of the stories about him and his sons are verging on the stuff of Saga. My own feeling is that it's a mistake by the likes of FTDNA to identify M222 solely with Niall (pronounced Neil in Irish as not Nile (anglisced) by the way). We see plenty of M222 show up in surnames linked to Connachta (Uí Bhriúin and Uí Fhiachrach "dynasties") who claim descent from his half-brothers, this would point to me that M222 is specifically a marker connected to wider Dál gCuinn (Dál Cuinn in old Irish).

The word Dál is sometimes translated as meaning "Share of" eg. "Share of ancestor", Dál Cuinn = Share of Conn, though in eDIL it's down more as sept/branch/division/"land held by tribe", so for example: (Screenshot I took to preserve formatting)

http://compsoc.nuigalway.ie/~dubhthach/Connachta-edil.png

Conn is probably mythical, some have argued that he's potentially a humanised deity, the pseudo-history syncrhonise his reign with that of Marcus Aurelius (mid 2nd century). His supposed grandfather is Túathal Techtmar, whose saga talks about a rebellion of the Province Kings overthrown his father, and that his pregnant mother (Daughter of the King of Alba -- Scotland, but in this case probably meaning Britain/Albion) fled back to her father. Túathal Techtmar then returned as a grown man from Britain to reclaim his rightful throne. This was during the 1st century by the way supposedly.

This saga if you ask me preserves a hint of movement/contact between Northern Britain and Ireland.

What's interesting with Chromo2 tests on M222 is that the more upstream branches seem more Scottish/Northern English (modern boundaries)

-Paul
(DF41+)

Dubhthach
01-21-2014, 04:07 PM
Here it is. In "Ancestral Journeys...", Jean Manco wrote,
"A more likely genetic signature of La Tene in Ireland is Y-DNA R1b-M222, carried by up to 44 per cent of men in parts of Northern Ireland today.
...
This is not the pattern we would expect from Irish migrants into Britain. So R1b-M222 hints at La Tene movements into Ireland."

I'd have to go back and read through all of that again, but I think the general idea is that M222 came into Ireland from Northern Britain with La Tene. Ironically, that would mean it came in as some kind of P Celtic group that integrated into a Q Celtic based Gaelic (to be) society.

Wasn't Niall supposed to come from the Connachta, which would be westerly in Ireland? The idea of a La Tene immigration seems to be at odds with that. Perhaps I have my Irish stories mixed up.

Mike,

It's worth pointing out that the Q/P differentiation in Insular Celtic at this stage would have been quite minor. Probably on order of difference between modern Scandinavian languages -- or a push between Dutch and German. It wouldn't be too hard to switch, these were closely related languages after all.

It's worth pointing out that the name of province of Connacht derives from fact that it was conquered by the Connachta. Even to this day people from province of Connacht are called "Connachta" in Irish language (been from Galway I'd qualify), in this case though it's geographic term as oppose to dynastical/tribal.

Before that it was called "Cóiced Ol nEchmacht" (Cóiced = Cúige eg. 1/5th eg. Province), the "Ol nEchmacht" coming from the tribal name of "Fír Ol nEchmacht" (Fír = men of)

Jean talks about La Tene movement, but alot of La Tene items are more specific into North-East Ulster (Ulaid) and date from 200BC onwards, perhaps with "Connachta" we are seeing movement more around the time of the Roman invasion of Britain perhaps tied to likes of Brigantes etc. (purely my own suggestion)

-Paul
(DF41+)

Dubhthach
01-21-2014, 04:10 PM
A fairly accessible "intro" can be found in this section on a book about Brian Boru:

http://books.google.ie/books?id=PH5wuxG7XOIC&lpg=PA27&ots=ZibP4uEVBh&dq=d%C3%A1l%20gCuinn&pg=PA27#v=onepage&q&f=false

jdean
01-21-2014, 04:23 PM
His closest match (4 off at 37 loci) is to a fellow called Hiles who's MDKA was a Heil from Frankfurt born 1710.

However his MDKA was born not 1/2 a mile from a place called Hilley in Hungary, anybody know enough about Hungarian names to know if this could be the source of his surname ?

TigerMW
01-21-2014, 04:34 PM
Good catch.

Must be this this guy, right?
17635 Hiles John (Johan) Heil, abt 1710, Frankfurt, Germany

http://www.familytreedna.com/public/Hiles/default.aspx?section=yresults

I see 17635 Hiles is 385=12,14 rather than 13,14 for 286325 Heil. Both are unusual but this suggests the mutation to 13 might be a second, later mutation and a red herring in the ancestral description of this group. 12,14 might be a better signature to search for for.

There is also this guy in the Hiles project who is 385=12,14.
46432 Hiles, b. Unknown Origin

So is this guy, but his 464 values are different so we have to be careful about these shorter haplotypes.
246586 Skiles, b. Unknown Origin

rossa
01-21-2014, 04:54 PM
What's interesting with Chromo2 tests on M222 is that the more upstream branches seem more Scottish/Northern English (modern boundaries)

-Paul
(DF41+)

Is it too early to tell if there are many examples of Irish people (with an assumed long standing Irish male lineage) who have these upstream snp's?

Sorry to go off topic (I couldn't find the thread on Worldfamiles forum) but what's the story with M222 and the O'Neill surname? From what I understand it doesn't turn up as much as you'd expect and it's possibly down to them getting almost wiped out after being kicked out of Aileach in Inishown.

Dubhthach
01-21-2014, 05:08 PM
Is it too early to tell if there are many examples of Irish people (with an assumed long standing Irish male lineage) who have these upstream snp's?

Sorry to go off topic (I couldn't find the thread on Worldfamiles forum) but what's the story with M222 and the O'Neill surname? From what I understand it doesn't turn up as much as you'd expect and it's possibly down to them getting almost wiped out after being kicked out of Aileach in Inishown.

Check out the article from Jogg (co-authored by the late John McLaughlin)
http://www.jogg.info/22/ONeill.pdf

It looks like the "O'Neill Variety" they describe is actually DF27+ (Z196-).

A basic glossary for those not familiar with early medieval Irish history:


Uí = plural eg. grandsons/descendants
Ua = singular eg. grandson/descendant, in modern Irish: Ó
Uí Néill = dynastical term for "descendants" of Niall Noígiallach (Niall of Nine Hostages)
Northern Uí Néill = Kingdom of Aileach (In Tuasiceart = The North), divided into three "Kindreds"
Southern Uí Néill = Midhe (Mí) eg. Meath -- divided into two "Kindreds"
Ua Néill/Ó Néill = surname, descendants of Niall Glúndub (Niall of the Black Knee) HighKing of Ireland 916-919AD, member of "Northern Uí Néill" dynasty


Within the concept of the Highkingship (Ard Rí) it rotated between "Dynastical branches" eg. Northern -> Southern -> Northern. Niall Glúndubh (member of Cinéal nEoghain of Northern Uí Néill) was succeded by Donnchadh Donn mac Flainn (King of Meath, member of Clann Cholmáin of Southern Uí Néill) etc. etc.

In case of the O'Neill family (Ua/Ó Néill) they were dispossed of Kingship of Aileach (and thus Cineal nEoghain) for over a century by their kinsmen the MacLochlainn (McLoughlin). It's possible that a NPE thus occurred in lineage, there's been some debate in academia that the lineage of the Ua/Ó Néill family is suspect during this period (mid/late 11th-12th centuries)

-Paul
(DF41+)

TigerMW
01-21-2014, 05:50 PM
Here is the line up of DF23+ Z2961- confirmed and suspected people. You could think of them, DF23*, as first cousins to M222.

f230189 McComb 49-23*- uas Ireland
f226516 Anglin 49-23*- uas Ireland
f50620 Anglin 49-23*- uas Ireland
f67089 Anglin 49-23*- uas Ireland
f63595 Anglin 49-23*- uas Ireland, Munster, Co. Cork
f35714 Elder 49-23*- uas Scotland
fN116876 Fogg 49-23*- uas zzzUnkOrigin
f286325 Heil 49-23*- uas Hungary, Southern Transdanubia, Baranya
f32773 Jenkins 49-23*- uas UK
fN115560 Joyce 49-23*- uas zzzUnkOrigin
f78065 Lamphier 49-23*- uas France, Languedoc-Roussillon

fN46295 Bonnet 49-23*-11-A Italy, Piedmont, Montoulles, Chambonsn (Waldensian French community)
f18917 Brun 49-23*-11-A France, Poitou-Charentes
f164101 Dougherty 49-23*-11-A Ireland
f165344 Greenlee 49-23*-11-A Scotland, Strathclyde, Argyllshire, Campbeltown
f161264 Johnston 49-23*-11-A zzzUnkOrigin
f97610 Kehoe 49-23*-11-A Ireland, Leinster, Co. Wexford

fN54996 Bennett 49-23*-152123 UK
f5155 Boun 49-23*-152123 England, East Midlands, Derbyshire, Bakewell
f176037 Davis 49-23*-152123 zzzUnkOrigin
f193936 Davis 49-23*-152123 zzzUnkOrigin
fN16755 Jones 49-23*-152123 England, South East, Berkshire, Reading
f185246 McCraw 49-23*-152123 zzzUnkOrigin
f147788 Reis 49-23*-152123 Germany
f14273 Roberts 49-23*-152123 Wales
f159206 zzzUnknown 49-23*-152123 zzzUnkOrigin

f69336 Bradley 49-23*-15-A England, South West, Somersetshire
f162881 Carroll 49-23*-15-A Ireland, Munster, Co. Tipperary, Ballingarry
f191293 Davis 49-23*-15-A zzzUnkOrigin
f11435 Elliott 49-23*-15-A Ireland, Ulster, Co. Donegal
f51058 Lewis 49-23*-15-A UK

f162597 Anderson 49-23*-15-WSW zzzUnkOrigin
f141902 Bryan 49-23*-15-WSW Wales
f90439 Davis 49-23*-15-WSW England
f239359 Davis 49-23*-15-WSW zzzUnkOrigin
f69464 Davis 49-23*-15-WSW zzzUnkOrigin
f156191 Evans 49-23*-15-WSW Wales, West, Cardiganshire
f81103 Griffin 49-23*-15-WSW Wales
f189010 Griffith 49-23*-15-WSW Wales
f5146 Hooper 49-23*-15-WSW England
f5140 Hooper 49-23*-15-WSW zzzUnkOrigin
f155519 LaMont 49-23*-15-WSW zzzUnkOrigin
f77238 Morgan 49-23*-15-WSW UK
f137289 Morgan 49-23*-15-WSW zzzUnkOrigin
f208316 Newton 49-23*-15-WSW zzzUnkOrigin
f133436 Pool 49-23*-15-WSW Wales
f157973 Pool 49-23*-15-WSW Wales
fN29376 Powell 49-23*-15-WSW Wales, Mid, Powys, Radnorshire, Lowes
f98110 Pugh 49-23*-15-WSW Wales
f113011 Scott 49-23*-15-WSW zzzUnkOrigin
f211712 Stephens 49-23*-15-WSW England
f19920 Stephens 49-23*-15-WSW Scotland
f249068 Stephens 49-23*-15-WSW zzzUnkOrigin
f244975 Stephens 49-23*-15-WSW zzzUnkOrigin
f197585 Stephens 49-23*-15-WSW zzzUnkOrigin
f156257 Stephens 49-23*-15-WSW zzzUnkOrigin
f110031 Stephens 49-23*-15-WSW zzzUnkOrigin
f25980 Stephens 49-23*-15-WSW zzzUnkOrigin
f26232 Stephens 49-23*-15-WSW zzzUnkOrigin
f120942 Stevens 49-23*-15-WSW Wales
f27039 Stevens 49-23*-15-WSW zzzUnkOrigin
f156124 Stevens 49-23*-15-WSW zzzUnkOrigin
f147036 Vaughan 49-23*-15-WSW Wales
f154440 Vaughan 49-23*-15-WSW Wales
f49637 Vaughan 49-23*-15-WSW Wales
f125079 Vaughan 49-23*-15-WSW Wales
f64716 Vaughan 49-23*-15-WSW zzzUnkOrigin
f108697 Vaughn 49-23*-15-WSW Wales
f146350 Vaughn 49-23*-15-WSW zzzUnkOrigin
f34742 Workman(Vaughan) 49-23*-15-WSW zzzUnkOrigin
f66192 zzzUnk(Stephens) 49-23*-15-WSW England

f41311 Eaton 49-23*-1922-A England, South East, Kent
f174499 Harvey 49-23*-1922-A England, South West, Somerset, South Somerset, Ashill
f99593 Taylor 49-23*-1922-A England

f137235 Caldwell 49-23*-1922-C Ireland
fN26081 Caldwell 49-23*-1922-C Scotland
f15940 Caldwell 49-23*-1922-C Scotland, Strathclyde, Lanarkshire, Glasgow
f123987 Caldwell 49-23*-1922-C zzzUnkOrigin
f79710 Carll 49-23*-1922-C UK
f36132 Kennedy 49-23*-1922-C zzzUnkOrigin
f39249 MacKenzie 49-23*-1922-C Scotland, Highland, Ross & Cromarty, Achiltibuie
f254860 McKenzie 49-23*-1922-C Scotland
f36459 McKenzie 49-23*-1922-C Scotland, Highland, Ross & Cromarty, Achiltibuie
f281207 McKinley 49-23*-1922-C zzzUnkOrigin
f192660 Merrill 49-23*-1922-C England
f82717 Taylor 49-23*-1922-C zzzUnkOrigin
f46661 Wilson 49-23*-1922-C Ireland, Ulster

f245296 Oliver 49-23*-39314-A zzzUnkOrigin
f259569 Oliver 49-23*-39314-A zzzUnkOrigin
fN115349 Vann 49-23*-39314-A zzzUnkOrigin

f13026 Davis 49-23*-44413 England, London
f11021 Davis 49-23*-44413 England, London, Middlesex
f216398 Davis 49-23*-44413 UK
f167247 Davis 49-23*-44413 Wales
f154875 Davis 49-23*-44413 Wales
fN2315 Davis 49-23*-44413 Wales
f116686 Davis 49-23*-44413 zzzUnkOrigin
f61733 Davis 49-23*-44413 zzzUnkOrigin
f255710 MacKenzie 49-23*-44413 Scotland
f119874 Stanton? 49-23*-44413 zzzUnkOrigin

f91774 Gray 49-23*-55715 Scotland
f86147 Gray 49-23*-55715 zzzUnkOrigin
f29258 Payne 49-23*-55715 UK
f177202 Ricks 49-23*-55715 zzzUnkOrigin

TigerMW
01-21-2014, 05:59 PM
Here is the line up of Z2961+ M222- confirmed and suspected people. You could think of them, Z2961*, as brothers to M222 or one step closer to M222 versus DF23*.

f175660 De Grey 49-2329-11-HyM England
f79367 Kelly(Ui Maine) 49-2329-11-HyM Ireland
f100219 Kelly(Ui Maine) 49-2329-11-HyM Ireland, Connacht, Co. Galway, Raheen, Gort
f162015 Kelly(Ui Maine) 49-2329-11-HyM Ireland, Connacht, Co. Roscommon
f291983 Kelly(UiMaine) 49-2329-11-HyM Ireland
f259127 Kelly(UiMaine) 49-2329-11-HyM Ireland
f267242 Kelly(UiMaine) 49-2329-11-HyM Ireland
f299488 Kelly(UiMaine) 49-2329-11-HyM Ireland
f265425 Kelly(UiMaine) 49-2329-11-HyM Ireland
f84929 Kelly(UiMaine) 49-2329-11-HyM Ireland, Connacht, Co. Galway
f151757 Kelly(UiMaine) 49-2329-11-HyM Ireland, Connacht, Co. Galway
f107869 Kelly(UiMaine) 49-2329-11-HyM Ireland, Leinster, Co. Laois
f289753 Kelly(UiMaine) 49-2329-11-HyM Ireland, Leinster, Co. Longford
f292015 Kelly(UiMaine) 49-2329-11-HyM zzzUnkOrigin
f160027 Madden 49-2329-11-HyM Ireland, Munster, Co. Clare (?Limerick)
f84927 O'Ceallaigh(UiMaine) 49-2329-11-HyM Ireland
f227854 O'Kelly(UiMaine) 49-2329-11-HyM Ireland
f84928 O'Kelly(UiMaine) 49-2329-11-HyM Ireland
fN33146 O'Kelly(UiMaine) 49-2329-11-HyM Ireland
f60472 Pugh 49-2329-11-HyM Wales
f128257 Pugh 49-2329-11-HyM Wales
f108037 Pugh 49-2329-11-HyM Wales
f122371 Pugh 49-2329-11-HyM Wales
f189772 Pugh 49-2329-11-HyM Wales, North, Gwyneed
f185074 Pugh 49-2329-11-HyM zzzUnkOrigin
f27822 Shannon 49-2329-11-HyM Ireland
f14875 Thomas 49-2329-11-HyM Wales, South Carmarthenshire, Llandybie
f159039 Trainor 49-2329-11-HyM Ireland
f108030 Trainor 49-2329-11-HyM Ireland
f81092 Traynor 49-2329-11-HyM Ireland
f132540 Traynor 49-2329-11-HyM Ireland, Ulster, Co. Cavan
fN5677 Traynor 49-2329-11-HyM Ireland, Ulster, Co. Tyrone, Dromore, Corlaghdergan
f230620 Walsh 49-2329-11-HyM England, North West, Greater Manchester, Manchester

fN22529 Davies 49-2329-1640-A Wales, Mid, Montgomeryshire, Llanfyllin
f74975 Humphrey 49-2329-1640-A Wales, North, Gwynedd, Caernarvonshire, Lleyn
fN1871 Warren 49-2329-1640-A UK

f225795 Allen 49-2329-1640-B Ireland, Leinster, Co. Westmeath, Rathconrath
f185294 Davies 49-2329-1640-B Wales
f178560 Davis 49-2329-1640-B Wales
f275300 Joyce 49-2329-1640-B Ireland
fN108400 Joyce 49-2329-1640-B Ireland, Connacht, Co. Galway
f250714 Lewis 49-2329-1640-B Wales
f144928 Morgan 49-2329-1640-B zzzUnkOrigin
f111218 Nallen 49-2329-1640-B Ireland
f74772 Phillips 49-2329-1640-B England
f160476 Welsh 49-2329-1640-B Ireland

f101343 Fancher 49-2329-1922-J England
f41836 Fancher 49-2329-1922-J England, London
f259442 Fanshaw 49-2329-1922-J England
fN10959 Johnson 49-2329-1922-J Ireland, Ulster, Co. Antrim, Belfast
f208254 Johnson 49-2329-1922-J UK
f60575 Johnson 49-2329-1922-J zzzUnkOrigin
f202301 Johnson 49-2329-1922-J zzzUnkOrigin
f208313 Johnson 49-2329-1922-J zzzUnkOrigin
f241592 Johnston 49-2329-1922-J Ireland
f262468 Johnston 49-2329-1922-J Ireland
f209861 Johnston 49-2329-1922-J Ireland
f68863 Johnston 49-2329-1922-J Ireland
f219125 Johnston 49-2329-1922-J Ireland
f117783 Johnston 49-2329-1922-J Ireland, Ulster
f95659 Johnston 49-2329-1922-J Ireland, Ulster, Co. Antrim, Aghnadore Townland
f90660 Johnston 49-2329-1922-J Ireland, Ulster, Co. Antrim, Londonderrykeighan
f231239 Johnston 49-2329-1922-J Scotland
f37464 Johnston 49-2329-1922-J Scotland
f185899 Johnston 49-2329-1922-J zzzUnkOrigin
f94865 Johnston 49-2329-1922-J zzzUnkOrigin
f224025 Johnstone 49-2329-1922-J Scotland, Dumfries and Galloway, Lockerbie
f73834 Landon 49-2329-1922-J France
f213191 Lewis 49-2329-1922-J zzzUnkOrigin
f207798 Martin 49-2329-1922-J Ireland, Ulster, Co. Antrim
f309425 Martin 49-2329-1922-J Scotland
f262153 Martin 49-2329-1922-J zzzUnkOrigin
f104400 Williams 49-2329-1922-J zzzUnkOrigin
f192062 zzzUnk(Sharp) 49-2329-1922-J zzzUnkOrigin

Heber
01-21-2014, 06:06 PM
I recently did an analysis of German L21.
Here are the results.

1249
1250
1251

Notice Heiland (Heil) which is similar to the Irish Hyland.
from the Anglicized form of Gaelic Ó hAoileáin, a variant of Ó Faoláin (see Whelan).

It is similar to the distribution of L21 SNPs in Ireland.
1252

People underestimate the exchanges between Ireland and Germany from 800AD - 1200AD.

http://www.pinterest.com/gerardcorcoran/irish-celtic-monastic-movement/
and 17th - 19th century
http://www.pinterest.com/gerardcorcoran/irish-wild-geese/

In my own family research I discovered hundreds of students travelling from Ross Carbery to its sister school Wurzburg and Regensburg in Medieval times.

http://www.rosscarbery.ie/view_event.php?EventID=175
http://fmrsi.wordpress.com/2013/06/18/event-rosscarbery-summer-school-22-23-june-2013/

TigerMW
01-21-2014, 06:17 PM
One last piece of information on the phylogeny above and around M222.
BISDNA (Jim Wilson) has found an SNP, S7073, that almost all M222+ people appear to fall under. In reality we should probably call this behemoth subclade S7073 rather than M222. There has been only one M222+ S7053- man found so far.

On Dec 17, 2013, probably based on info from BISDNA

I have been told that the paper trail of the sole S7073- man is late 1700s in Fife but I still don't have a surname. Given that there were Kennedy families in Dunfermline by the mid 1500s (not to mention some individual males a century earlier) its hard to say how valuable this piece of information is. http://archiver.rootsweb.ancestry.com/th/read/DNA-R1B1C7/2013-12/1387266694

Jon
01-21-2014, 07:02 PM
Gerard, my father has done quite a bit of research into Scoto-Irish religious movement into Germany and Austria. It's amazing how many 'Schottenstifte' (Scottish Monasteries) there were. Except they weren't strictly 'Scottish': they seem to have been a mix of Irish and Scots, in the early days of the Celtic monastic orders, where these terms basically were one and the same. The Schottenstift in Vienna was taken back by the Viennese after complaints of the monks 'drinking strong spirits and playing games with balls'. Could be Scots or Irish, I'd say!! Regensburg was another that even remained 'Scottish' until much later. This is certainly a major historical element, which surely must have led to the spread of some of these 'celtic' haplogroups (back?) into Germany/Austria.

Jon
01-21-2014, 07:04 PM
PS...Gerard, it strikes me that you are based in Vienna, right? If so, you probably know all that already :)

Heber
01-21-2014, 07:21 PM
Jon,
I lived and worked in Munich last year and spent my weekends visiting over 50+ Schottenstift in Bavaria, Austria, Switzerland, Czech. many are now UNESCO Heritage sites eg. St. Gallen., Regensburg, Wurzburg.
The German word for Irish was Schotten and in later years they were a mixture of Irish and Scottish so I am sure L21 is well represented. What is extraordinary is visiting St Gallen World Heritage Library and seeing the Gaelic Insular script from 800 - 1200 transforming to a Gothic script thereafter.
http://en.wikipedia.org/wiki/Hiberno-Scottish_mission

Jon
01-21-2014, 07:36 PM
Fantastic: I've been to a few myself, they are stunning. The religious aspect, also of irish monastic settlement in Scotland in the time of the conversion of the Picts etc., surely must have been an additional aspect to the usual 'war band' theories. The Celtic church had real clout it seems. Thanks for the link.

Heber
01-21-2014, 07:48 PM
Jon,
If you look in the Abbey Graveyard of Iona you will find the remains of 50+ kings of Scotland, Ireland and Norway although their graves are unmarked because of erosion. This would make a great ancient DNA project (probably L21) if only we could get the right permits. So yes the Celtic church had great influence in shaping the history of Scotland and possibly Europe. They certainly played a leading role in the Carolingian Empire.

TigerMW
01-21-2014, 07:56 PM
....
The German word for Irish was Schotten and in later years they were a mixture of Irish and Scottish so I am sure L21 is well represented. ...

I'm sure there are some folks from the Isles that have traveled into Germany during the historic period, but how do we explain a DF23* person who doesn't fit any Isles clusters?

Heber
01-21-2014, 08:28 PM
I'm sure there are some folks from the Isles that have traveled into Germany during the historic period, but how do we explain a DF23* person who doesn't fit any Isles clusters?

Mike,

I'm sure there are some folks from the Isles that traveled into Europe prior to the historic period.:). Cunliffe has demonstrated this clearly in Britain Begins. I will look closer at the DF23* case. Do you have his id. I hope our imminent Tsunami of SNPs will clarify a lot of these cases.

Edit: I looked at your list above. Are you referring to Heil. I would guess this is a version of Heiland (Hyland) AKA Whalen. Most of these names appear to be HyMany or Connachta and likely to be DF49. In any event they are DF13, the majority appear to be Gaelic.

Jon
01-21-2014, 08:52 PM
Incidentally, do any of you know what the situation is regarding legals rights to extracting DNA samples from ancient burial sites? Presumably a skeleton dug up during an archaeological excavation is OK, but I guess on Iona the site still counts as an established burial ground..?

Dubhthach
01-21-2014, 09:08 PM
Given the Acidic nature of soils in Iona and general area you probably wouldn't even find Skeletons let alone ancient-DNA. Just the same way as the body wasn't found in Sutton Hoo, it had been gradually eaten away by the acidic sandy soil

-Paul
(DF41+)

Heber
01-21-2014, 09:22 PM
Given the Acidic nature of soils in Iona and general area you probably wouldn't even find Skeletons let alone ancient-DNA. Just the same way as the body wasn't found in Sutton Hoo, it had been gradually eaten away by the acidic sandy soil

-Paul
(DF41+)

Paul,

Hopefully they were buried in stone graves with cap stones and better preserved.

http://www.englishmonarchs.co.uk/macalpin_3.htm

"The Abbey was to become the mausoleum of the early Celtic Kings of Scotland who wished to be buried near to St. Columba. Their burial ground, known as Reilig Odhrain and dedicated to St. Oran, lies to the south-west of the priory at Iona. An inventory of 1549 recorded 48 Scottish, 8 Norwegian and 4 Irish kings buried there. The inscriptions are said to have worn away by the end of the seventeenth century."

If we can excavate the bones of Richard III and lay him to rest again with dignity in a Cathedral surely we should be able to do that for the Kings of Scotland, Ireland and Norway.

TigerMW
01-21-2014, 09:25 PM
Mike,

I'm sure there are some folks from the Isles that traveled into Europe prior to the historic period.:). Cunliffe has demonstrated this clearly in Britain Begins. I will look closer at the DF23* case. Do you have his id. I hope our imminent Tsunami of SNPs will clarify a lot of these cases.
I agree. There were probably prehistoric migrations in both directions (to and from the Isles) so anything is possible. I think the larger population movements prehistorically and even up through the Roman Era were from the continent to the Isles though. I find the Wessex culture trading links to the Middle Rhine and Unetice folks quite interesting, although that may have been before DF23's time.


Edit: I looked at your list above. Are you referring to Heil. I would guess this is a version of Heiland (Hyland) AKA Whalen. Most of these names appear to be HyMany or Connachta and likely to be DF49. In any event they are DF13, the majority appear to be Gaelic.

Yes, Heil. I think you are saying the surnames appear to be Gaelic?

This becomes difficult to ascertain based on surname. Ancestry.com claims it has a German origin.

"German: from a pet form of Heinrich.Dutch and North German: from a short form of the Germanic female personal name Heila, derived from hail ‘whole’"
http://www.ancestry.com/name-origin?surname=heil

Heber
01-21-2014, 09:30 PM
Yes, Heil. I think you are saying the surnames appear to be Gaelic? ... so then Heil is a derivative of a Gaelic name or word?

This becomes difficult to ascertain based on surname. Ancestry.com claims it has a German origin.

"German: from a pet form of Heinrich.Dutch and North German: from a short form of the Germanic female personal name Heila, derived from hail ‘whole’"
http://www.ancestry.com/name-origin?surname=heil

Mike,

1253

Heiland is one of the few German L21 results so it is possibly related to that. We cannot be sure until we have get better SNP resolution.

Jon
01-21-2014, 09:53 PM
Don't want to overly complicate matters, but I've heard Heil in a religious context, from the German 'heilig' meaning holy. The surname Heilig also in fact exists, and is supposed to refer to one who was associated with a religious order, in much the same way as MacMillan (angl. of 'son of the tonsured one') in Scotland. Could be pure and utter coincidence, and indeed probably is. But the possibility is intriguing...

Dubhthach
01-21-2014, 10:29 PM
in much the same way as MacMillan (angl. of 'son of the tonsured one') in Scotland.

That's a too literal translation, it would be akin to saying that somone called Peterson was "Son of the Rock".



Mac MAOLÁIN—IV—M'Mowllane, M'Moylan, MacMullan, MacMullen, MacMullin, MacMullon, MacMillan, MacMillen, MacBlain, Mullin, Mullins, &c.; 'son of Maolán' (diminutive of maol, bald). In the 12th century, Mac Maoláin was lord of Gaileang Breagh, in the north of the present Co. Dublin, but in later ages the name has been confined to North-East Ulster. There is also a Scottish Mac Maoláin.


In reality the name means "son of a man called Maolán". Direct translation would be "little baldy", just way "Paul" is derived from latin adjective for "small"/"few" -- I'm hardly small so I don't think we can take it for granted that Maolán was either "small and bald" or tonsured!

-Paul
(DF41+)

newtoboard
01-21-2014, 10:39 PM
Did La Tene leave any genetic influence in Turkey and the Balkans? I know it was supposed to have had a significant cultural impact n that area.

Eochaidh
01-21-2014, 11:24 PM
Over on that thread about Jean's book, you mentioned her idea that M222 could represent the arrival in the Isles of some continental settlers bearing La Tene culture and skills. Maybe that's right, or it began with DF23.
Batar trí prímcheinéla i nHére, .i. Féini 7 Ulaith 7 Gáilni .i. Laigin.
"There were three principal kinships in Ireland: the Féini, the Ulaidh, and the Gáilni, i.e., the Laighin."
-- From an Eighth Century legal tract.

All of the La Tène artifacts unearthed in Ireland have been found in the northern part. The metal finds are generally made from bronze, but in the La Tène style. Other finds are the La Tène Beehive querns found in the same general areas, but not in the exact locations. The metal objects are found in the good land and the querns are found in the poorer land. It is believed that this indicates a class distinction and indicates a population change beyond just mounted warriors.

The three principle tribes in this area were the Connachta(Féini), Ulaidh, and the Laighin and it was about these peoples that the Ulster Cycle and in particular, the Táin Bó Cúailnge were composed.

Pronunciations:
Ulaidh = Uly; Tir Ulaidh => Ulster
Laighin = Layin; Tir Laighin => Leinster
Emhain Macha = Avan Macha
Cruachain = Cruacin
Táin Bó Cúailnge = Tawn Bo Cooley; Cattle driving of Cooley - Cooley is a peninsula in County Louth.

1254

Work has been done in the last few decades by archaeologists who have identified two clear horizons which unite the Connachta, Ulaidh, and the Laighin by way of what are called their Royal Sites which are identified in the Táin.

These are the 'Royal Sites' of the Táin Bó Cúailnge which have been identified and studied. Emhain is the best studied and Cruachain the worst, but that is improving.
Knockaulin - Dún Ailinne of the Laighin
Tara - the most important
Navan Fort - Emhain Macha of the Ulaidh
Rathcroghan - Cruachain of the Connachta

http://i1125.photobucket.com/albums/l589/Knockbridge1/DNA/figure_of_eight_lynn500.png

Horizons (summarized):
4th century BCE - late second BCE: A series of figure-of-eight structures are built in Emhain, Dún Ailinne and Tara. (Cruachain has the top half, i.e. the large circle and the east facing walls). They have no parallel anywhere in Europe.
1st century BCE (95 BCE at Emhain by dendrochronology): The figure-of-eight structures are removed and a "40 meter structure" is built at all 4 sites. They have no parallel anywhere in Europe.

It seems that the archaeology reveals a direct prehistoric datable connection among the Connachta, Ulaidh, and the Laighin which parallels the stories of the Ulster Cycle. This may mean something about the spread of La Tène and possibly M222.

alan
01-21-2014, 11:43 PM
It may seem that way but in fact Connaught after Ulster is the most blessed with La Tene material of the 4 Irish provinces. Also back in the times of the Ulster Cycle the Connaught army seem to have featured a crack force of the Fir Domnainn or Dumnonii. In Ireland that tribe have associations with both Leinster and Connaught.

I think in Mallory's book he noted someone else's theory approvingly that there was probably a mi of Q and P Celtic tribes in Ireland and that Ptolemy's map suggests both types. He suggested that, as a very conservative group who had the unique privilege of being protected when crossing tribal borders and were the only unifying force in a very divided country, it may have been the sacred classes like druids etc who ran the large assemblies in Iron Age Ireland who kept the Q Celtic on top. This fits the archaeological evidence which shows Ireland was a peculiarly ritualistic country at the time with totally unique enormous ritual/royal/assembly sites which were way beyond anything known elsewhere in Europe. The enormous investment in these sort of sites is in stark contrast to the lack of houses, pottery, coinage, rich burials, significant foreign trade etc recovered from this time in Ireland. It seems that Ireland was in a unique position for its language trends to be dictated by the sacred class, including the druids. I am almost convinced that this is the real reason why Q came to be the chosen language when in fact there had been a time when there was a mix of P and Q. A continental hint of the preference for archaic Q forms by the Druids is contained in the Coligny Calender from France which includes Q forms of words.


Here it is. In "Ancestral Journeys...", Jean Manco wrote,
"A more likely genetic signature of La Tene in Ireland is Y-DNA R1b-M222, carried by up to 44 per cent of men in parts of Northern Ireland today.
...
This is not the pattern we would expect from Irish migrants into Britain. So R1b-M222 hints at La Tene movements into Ireland."

I'd have to go back and read through all of that again, but I think the general idea is that M222 came into Ireland from Northern Britain with La Tene. Ironically, that would mean it came in as some kind of P Celtic group that integrated into a Q Celtic based Gaelic (to be) society.

Wasn't Niall supposed to come from the Connachta, which would be westerly in Ireland? The idea of a La Tene immigration seems to be at odds with that. Perhaps I have my Irish stories mixed up.

JRW
01-22-2014, 12:01 AM
If my family's history is in anyway typical for one of German ethnicity, assuming historic origins from a single 19th century data-point could be very misleading. My g4 grandfather, who was born in Göttingen, Hanover, 1754, had children and/or grandchildren who moved to and resided in towns/cities that are now in present-day Slovakia, Hungary, Poland, United Kingdom, Germany, as well as Mexico, Canada and the U.S.. His father, in turn, was from a different town and state in "Germany," Jena in Thüringen, and his paternal grandfather yet from another (Würzburg, Franconia). Perhaps my views are overly influenced by my own family's mobility, but I think a German surname (e.g., Heil) would be strongly indicative of German ethnicity -- regardless of where it may be found (of course, there are always exceptions).

I also caution about discounting L21's and its older subclades' ancient relationship with the area of present day Germany. Although there are definitely German men who descend from Isles ancestors (and I think these are easily spotted - either through surname -- e.g., Schott -- or GD relationships), there are many more L21 men of German ethnicity who, given their GDs to those in the Isles, would not appear to have the potential for a historic period Isles-to-continent migration in their ancestry. I may be in the minority on this, but, IMHO, I think L21, including some of its subclades, constituted a significant portion of the population among the tribes residing along the North Sea coast in what is now Belgium, the Netherlands, Germany and Denmark before the Völkerwanderung (migration period). The effect of the Völkerwanderung (and some earlier small migrations) was to send most of that L21 to the Isles (including Ireland), to Gaul, and up the Rhine. The reason why we haven't seen any "German" L21 STR clusters to date is that the remaining L21 became very diluted due to other population movements, wars, plagues, etc., thereby precluding material "founder effects."

rossa
01-22-2014, 12:21 AM
It may seem that way but in fact Connaught after Ulster is the most blessed with La Tene material of the 4 Irish provinces. Also back in the times of the Ulster Cycle the Connaught army seem to have featured a crack force of the Fir Domnainn or Dumnonii. In Ireland that tribe have associations with both Leinster and Connaught.

I think in Mallory's book he noted someone else's theory approvingly that there was probably a mi of Q and P Celtic tribes in Ireland and that Ptolemy's map suggests both types. He suggested that, as a very conservative group who had the unique privilege of being protected when crossing tribal borders and were the only unifying force in a very divided country, it may have been the sacred classes like druids etc who ran the large assemblies in Iron Age Ireland who kept the Q Celtic on top. This fits the archaeological evidence which shows Ireland was a peculiarly ritualistic country at the time with totally unique enormous ritual/royal/assembly sites which were way beyond anything known elsewhere in Europe. The enormous investment in these sort of sites is in stark contrast to the lack of houses, pottery, coinage, rich burials, significant foreign trade etc recovered from this time in Ireland. It seems that Ireland was in a unique position for its language trends to be dictated by the sacred class, including the druids. I am almost convinced that this is the real reason why Q came to be the chosen language when in fact there had been a time when there was a mix of P and Q. A continental hint of the preference for archaic Q forms by the Druids is contained in the Coligny Calender from France which includes Q forms of words.

If Ireland was so cut off for most of the Iron Age how did the M22 carriers come to dominate? Was it a case that the Irish groups military capacity had become diminished which made it easy for the British groups?

oneillabu
01-22-2014, 12:21 AM
I don't know where DF49 and DF23 originated and I don't know if M222 is really connected to Niall, ancient Irish King. Its a thought provoking conversation, though.

Here is some new information. One does not a trend make, but here is an interesting new DF23+ test result:

286325 Martin Heil, b.1877, Baranya, Southern Transdanubia, Hungary

I've seen him in the L21 project for some time but was always perplexed. I didn't see any SNP testing until now. My spreadsheets would pick him as a potential DF49* or DF23* type, but every time I would check GD's they'd be so far out that I'd discard his haplotype from the spreadsheet updates.

He's got Geno 2 results now, and sure enough, he is DF23+. I don't see an M222+ so I presume he is M222- but we should look at his raw results to be sure.
http://www.familytreedna.com/public/R-L21/default.aspx?section=ysnp

Hungary is a quite a ways from Ireland but we do have some cases of people traveling quite widely in our historic period. However, I think the other factor is the STR signature. He has a high DYS481 value and DYS413 mutations which are typical in M222 and some parts of DF23. However, the point is that the rest of his STRs scarely match any clusters.

His closest GDs to anyone within L21 are GD=15 @67 which means he broke away from the rest of DF23 a long time ago, perhaps about the time of Caesar.

His closest GD's (15) are:

f286325 Heil R1b-L21>DF13>DF49>S476>DF23 DF23+ L21+ L21+ Z290+ (G2) 49-23*- uas
f96218 Grenon zzL21suspect 49- uas (France)
f95659 Johnston zzL21suspect 49-2329-1922-J
f101298 MacGregor zzL21suspect z43811-10
f36459 McKenzie zzL21suspect 49-23*-1922-C

Grenon is just suspected L21 but he does match some of those signature markers for DF23 types, which is why I included him in the file.

Of course, the Heil surname could relate to German immigration east into Hungary.


There is clearly a connection between Thomas Lamphier, 1570 Lanquedoc, France (78065) who is also DF23+ and Martin Heil, if you compare the last eight markers of 67 you will see they match exactly even though the GD between them is over 20 at for 67 markers. This points to a French origin for DF23 however DF49 seems to be Brythonic in origin which is consistent with the forced migration to Brittany caused by the Saxon invasion of Britain.

I believe that there was a very large DF23 / M222 element amongst the Norman army of William the Conqueror and that a lot of Irish M222 was spread by the Normans, some of whom adopted Irish surnames. This would account for the large number of surname matches to M222 people within the timeframe of the adoption of surnames; some people have over a thousand 67 marker matches with completely unrelated surnames. This would also account for the very large amount of British M222, just look through the names in the DF49 project and it is obvious that there is no connection to the ancient Kings of Ireland there, even the Hy Maine names match to Welsh and Norman surnames such as Joyce who was one of the Norman tribes of Galway.

alan
01-22-2014, 12:25 AM
Eochaid raises a good point that again is relevant to Mallory's idea that the sacred class in Ireland may have set the linguistic standards. Despite a variable amount of La Tene material in Ireland, other aspects speak of surprising uniformity. That includes the royal sites - really ritual assembly sites would be a better term. There are now more known than the big 4 listed above and IMO they almost certainly were known across Ireland. They had many similarities which suggests a super-tribal conceptual guiding hand. In Iron Age Ireland that would have been Druids etc rather than Kings as the latter did not have authority out of their own territory and the idea of High King is really an invention of later times pushed by monks at Armagh and other Ui Neill friendy monasteries who wanted a secular reflection of their ecclesiastical ambitions to central power.

Other areas of uniformity include the burial patterns (rare, cremations, round barrow, not much grave goods etc), the use of linear defensive ditches, the rejection of domestic or burial pottery, a mode of living and building dwellings which left little trace either due to ephemeral nature or design that didnt leave many post holes in the subsoil etc.

In general the variation in quantity of La Tene material across Ireland is real but is overemphasised. A lot of the concentration in the north (and its the northern two-thirds rather than just Ulster) may be down to early collectors being located there. Lets put it another way, no 'alternative' tradition in metalwork or culture of a non-La Tene type (other than hangovers from the local Bronze Age) has ever been found. It has also been shown by recent monitoring of roads, pipelines etc that the whole island shared similar settlement types and similar periods of peak dates for building and the converse periods of lack of settlement. The only difference is the frequency of La Tene metalwork that is found. I think that is at least partly an illusion as there was simply no other show in town in terms of artistic showing off for the elite.

alan
01-22-2014, 12:41 AM
Its hard to answer that. I get the impression that politically Ireland was extremely fragmented and few kings in the Iron Age had much power other than very local. Its easy to see how a constant trickle of war bands could come in in a situation like that although due to climatic downturn, growth of bog and loss of importance of its metals, Ireland may not have been a very attractive distination in the Iron Age.

The Celtic peoples always had an unusual dual power of secular kings and the super-tribal wider power of the sacred druid class. In Ireland I think that simply reached an extreme with the sacred classes being the only source of power beyond local.Why were the Irish kings so weak in comparison to the sacred class compared to other Celtic areas? I think Ireland elites really took a mortal blow when Iron came in and ruined their long standing power based on copper and gold which had given them a lot of wealth in the Bronze Age. Ireland was an outstanding star of the Bronze Age and probably the elites had become too based on trade and metals. Other areas that never had such an advantageous position in metals probably had elites whose power was a little less dependent on this. Places like Lowland England for instance had always been down the metal chain/middlemen due to lack of metal sources and were much more used to being dependent and presumably had adapted to power being based more on land.

The huge ritual/ assembly sites were almost certainly built for assemblies and one of the main aspects of these gatherings would involve speaking. The druids probably controlled a lot of this. Only these huge ritual sites had the ability to influence beyond very local chiefdoms. They seem to have served large regions somewhat like the modern four provinces. The same Druidical class probably controlled or had a major role in all of the assembly sites. This put them and their assemblies in a unique position to dictate linguistic trends, social trends and OPINION. The power of druids and poets on public opinion was so great that they could destroy kings and this is features in early Irish literature. I think its also commented on in Gaul although they had much more serious secular powers to compromise with than would have been the case in Ireland.

Druids are widely thought to have practiced deliberate archaism and this could be why Q Celtic remained (or possibly revived and expanded). Ireland may have gone through a very long period from c. 700BC to perhaps 300AD or later when the druids had very little in the way of any sort of secular counterbalance. They may have only been challenged by contact with Roman Britain, materialism through raids and settlements there by the Irish, new Roman gods and ultimately Christianity.


If Ireland was so cut off for most of the Iron Age how did the M22 carriers come to dominate? Was it a case that the Irish groups military capacity had become diminished which made it easy for the British groups?

alan
01-22-2014, 12:55 AM
As for clades and their distribution across the isles, another theme of Mallory's book is that there was essentially throughout most of the Neolithic and Bronze Ages something of a superhighway along the Irish Sea that meant huge amounts of cultural aspects and latest trends were closely shared c. 4000-700BC between the island with much much less direct contact with the continent. The model that makes most sense would be a constant low level trickle movement with some peaks and trough throughout that period. So, if Ireland and Britain share a great deal of clades below L21 then this is totally in line with archaelogical expectation. There are so many things that Ireland and Britain share most closely with each other rather than with the continent that if one wanted to pinpoint an exact archaeological signal for a clade then we would be overwhelmed with options. This can only be refined and some sort of stab at interpretation carried out if clade dating become accurate and agreed on.

rms2
01-22-2014, 03:46 AM
One thing to remember about Irish and/or Scottish monasteries is that they were populated by monks who had taken vows of poverty, chastity, and obedience. In other words, the monks were supposed to be celibate. That is not to say they never lapsed and gave in to temptation, but the idea that they had a real, detectable genetic impact on the Continent seems to me to really strain credulity.

rms2
01-22-2014, 03:57 AM
If my family's history is in anyway typical for one of German ethnicity, assuming historic origins from a single 19th century data-point could be very misleading. My g4 grandfather, who was born in Göttingen, Hanover, 1754, had children and/or grandchildren who moved to and resided in towns/cities that are now in present-day Slovakia, Hungary, Poland, United Kingdom, Germany, as well as Mexico, Canada and the U.S.. His father, in turn, was from a different town and state in "Germany," Jena in Thüringen, and his paternal grandfather yet from another (Würzburg, Franconia). Perhaps my views are overly influenced by my own family's mobility, but I think a German surname (e.g., Heil) would be strongly indicative of German ethnicity -- regardless of where it may be found (of course, there are always exceptions).

I also caution about discounting L21's and its older subclades' ancient relationship with the area of present day Germany. Although there are definitely German men who descend from Isles ancestors (and I think these are easily spotted - either through surname -- e.g., Schott -- or GD relationships), there are many more L21 men of German ethnicity who, given their GDs to those in the Isles, would not appear to have the potential for a historic period Isles-to-continent migration in their ancestry. I may be in the minority on this, but, IMHO, I think L21, including some of its subclades, constituted a significant portion of the population among the tribes residing along the North Sea coast in what is now Belgium, the Netherlands, Germany and Denmark before the Völkerwanderung (migration period). The effect of the Völkerwanderung (and some earlier small migrations) was to send most of that L21 to the Isles (including Ireland), to Gaul, and up the Rhine. The reason why we haven't seen any "German" L21 STR clusters to date is that the remaining L21 became very diluted due to other population movements, wars, plagues, etc., thereby precluding material "founder effects."

The French archaeologist and linguist Henri Hubert agreed with your basic premise, but he has the Beaker Folk (whom I am taking the liberty of interpreting as a proxy for L21) departing the North Sea coast for Britain in the Bronze Age and leaving the place nearly deserted. Germanic tribes would begin to settle there in about 700 BC and would succeed in pushing the Celts west of the Rhine around 200 BC or so.

So, in my humble opinion, there wasn't much L21 among the Anglo-Saxons; they were mostly U106 and I-M253. L21 had retreated west of the Rhine into northern Gaul long before the Migration Period. That is why Myres et al and Busby et al found L21 at such low levels in Germany. Still, there is some there, representing descent from the Celts who once lived in western and southern Germany. I disagree with the attempts to chalk the L21 on the Continent up to men from the Isles.

TigerMW
01-22-2014, 12:42 PM
Did La Tene leave any genetic influence in Turkey and the Balkans? I know it was supposed to have had a significant cultural impact n that area.
I think that is a different thread, but it is a worthwhile point to examine for sure. If we somehow found that L21 was carried (not the only hg, just one of) with La Tene into Britain then we might look to see if it went eastward towards Anatolia too.

It doesn't look like L21 made it anywhere close to that far east so L21 provides no evidence of a La Tene movement eastward... if there is L21 connection to La Tene in the first place.

TigerMW
01-22-2014, 12:52 PM
...
I believe that there was a very large DF23 / M222 element amongst the Norman army of William the Conqueror and that a lot of Irish M222 was spread by the Normans, some of whom adopted Irish surnames. This would account for the large number of surname matches to M222 people within the timeframe of the adoption of surnames; some people have over a thousand 67 marker matches with completely unrelated surnames. This would also account for the very large amount of British M222, just look through the names in the DF49 project and it is obvious that there is no connection to the ancient Kings of Ireland there, even the Hy Maine names match to Welsh and Norman surnames such as Joyce who was one of the Norman tribes of Galway.

I hadn't thought of this perspective - DF23/M222 going with the Normans into England, Wales and then Ireland.

Let me do a little checking. There is a recent French study out.

EDIT/ADDITION:

Here is the study.

"Y-chromosomal DNA analysis in French male lineages" by Ramos-Luis, et. al., December, 2013.

The included L21 and broke out M222 separately as R1b1b2a2e.

They say they surveyed 558 Frenchmen from these locations.
Bretagne
Ile de France
Nord Pas de Calais
Alsace
Auvergne
Provence Alpes Cote dAzur
Midi Pyrenees

If I'm reading their percentages correctly, they only found one M222+ person. He was not from Bretagne, Ile de France or Nord Pas de Calias, but was from Midi Pyrenees.

That's just M222. I don't know, but it seems like a longshot for the DF23* in Wales to have gotten there via the Normans.

Keep in mind our DNA project data has many times more people from Ireland than from Wales. The Irish project has been 5000 and 6000 members. The Wales project has less than 1000 so when you see a heavy Welsh influence on a cluster/variety, as we see in parts of the DF23* varieties, the significance is more.

The breakout of DF23* varieties is back in reply #11 (click here.) (http://www.anthrogenica.com/showthread.php?1907-Origins-of-M222-DF23-the-NW-Irish-and-connections-to-Niall-of-the-Nine-Hostages&p=27922&viewfull=1#post27922). The Z2961* varieties, including the Ui Maine, are in reply #12 (click here.) (http://www.anthrogenica.com/showthread.php?1907-Origins-of-M222-DF23-the-NW-Irish-and-connections-to-Niall-of-the-Nine-Hostages&p=27925&viewfull=1#post27925)

Heber
01-22-2014, 02:24 PM
U
One thing to remember about Irish and/or Scottish monasteries is that they were populated by monks who had taken vows of poverty, chastity, and obedience. In other words, the monks were supposed to be celibate. That is not to say they never lapsed and gave in to temptation, but the idea that they had a real, detectable genetic impact on the Continent seems to me to really strain credulity.

Richard,

The network of Irish and Scottish and later Benedictine medieval monasteries, Schottenshift, formed the backbone of the Holy Roman Empire which was to last 1,000 years from 800 - 1800.
They comprised schools, universities, hospitals, hostels, farms, market gardens, bee keeping, brewing, scriptorium, astronomy, forges, stonework, metalworking, safe havens for travellers as well as churches, cathedrals, abbeys, convents and monasteries. Many of the valued manuscripts from 800 - 1200 were produced by these institutions. Up to 30 % of the surrounding land belonged to the monastery under the protection of the local kings or chieftains. They enjoyed the direct protection and sponsorship of most of the early Holy Roman Emperors. Populations in monastic settlements could reach five thousand with the majority lay or married. They formed the nucleus of the great Medieval cities such as Koln, Frankfurt, Wurzburg, Worms, Regensburg, St Gallen, Salzburg, Constance etc. Life expectation and ability to pass on genes was higher than that of the surrounding warring tribes and primitive agriculture. The monasteries brought with them a revolution in land organisation. There was constant recruitment of students and novices from sister houses in the Isles, some of whom entered the monastery, many did not.

Out of the 11K+ samples in the L21 database (thanks to the hard work of Mike and Richard), there are only a handful of confirmed L21 SNPs in Germany despite aggressive recruitment and these mirror the structure of L21 in the Isles. The few exceptions I have studied can be traced back to a Gaelic origin.

I believe there were several back migrations of L21 from the Isles in the past 2,000 years. The Irish diaspora is exceptional in having well over ten times the source population living abroad. I don't get the idea of L21 marching across Europe like Lemmings and stopping at the Atlantic.

Genographic who have the latest NGS Datasets from the Mayo and Asturias projects indicated that M222 could have originated in Mayo. This is Connachta country. We will not know for sure until the details are published and especially the imminent arrival of a wave of new SNPs. Even more enlightening will be a greatly expanded Phylogenetic Tree.

My understanding is that there were three major expansions of R1b. L2 under U152 associated with the Le Tene and Halstatt Iron Age movement dominating Gaul, Centred on Alesia and Halstatt. DF27 a Bronze Age movement with highest frequency in Iberia and the Atlantic facade. L21 a Bronze Age Atlantic Bell Beaker movement and later M222 with highest frequency in The Isles and Ireland and especially the west of Ireland. I have performed detailed analysis of all the L21 SNPs and they tell a similar story.

I have updated my analysis of L21 and downstream SNPs,
DF21, DF49, C4466, L513, Z253, Z255, L1335, DF41.
http://www.pinterest.com/gerardcorcoran/r1b-l21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df49/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-c4466/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l513/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z253/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z255/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l1335/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df41/

The greatly expanded Phylogenetic tree should also show the sequence of expansion.

If the data shows something different, I will be first to embrace it and adjust my understanding accordingly.

jdean
01-22-2014, 03:42 PM
I have updated my analysis of L21 and downstream SNPs,
DF21, DF49, C4466, L513, Z253, Z255, L1335, DF41.
http://www.pinterest.com/gerardcorcoran/r1b-l21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df49/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-c4466/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l513/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z253/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z255/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l1335/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df41/

The greatly expanded Phylogenetic tree should also show the sequence of expansion.

If the data shows something different, I will be first to embrace it and adjust my understanding accordingly.

Your analysis of DF49 is swamped with M222 data, DF49 is closer to P312 than M222

rms2
01-22-2014, 03:52 PM
U

Richard,

The network of Irish and Scottish and later Benedictine medieval monasteries, Schottenshift, formed the backbone of the Holy Roman Empire which was to last 1,000 years from 800 - 1800.
They comprised schools, universities, hospitals, hostels, farms, market gardens, bee keeping, brewing, scriptorium, astronomy, forges, stonework, metalworking, safe havens for travellers as well as churches, cathedrals, abbeys, convents and monasteries. Many of the valued manuscripts from 800 - 1200 were produced by these institutions. Up to 30 % of the surrounding land belonged to the monastery under the protection of the local kings or chieftains. They enjoyed the direct protection and sponsorship of most of the early Holy Roman Emperors. Populations in monastic settlements could reach five thousand with the majority lay or married. They formed the nucleus of the great Medieval cities such as Koln, Frankfurt, Wurzburg, Worms, Regensburg, St Gallen, Salzburg, Constance etc. Life expectation and ability to pass on genes was higher than that of the surrounding warring tribes and primitive agriculture. The monasteries brought with them a revolution in land organisation. There was constant recruitment of students and novices from sister houses in the Isles, some of whom entered the monastery, many did not.

It is doubtful that all those working in and around these monastic establishments were from the Isles. As I said, it strains credulity to attempt to attribute most of the L21 in Germany to them.



Out of the 11K+ samples in the L21 database (thanks to the hard work of Mike and Richard), there are only a handful of confirmed L21 SNPs in Germany despite aggressive recruitment and these mirror the structure of L21 in the Isles. The few exceptions I have studied can be traced back to a Gaelic origin.

What "aggressive recruitment"?

Besides that, we do not see German L21s displaying typical Isles haplotypes and membership in Isles clades. So, no, L21 SNPs in Germany and elsewhere on the Continent do not "mirror the structure of L21 in the Isles".



I believe there were several back migrations of L21 from the Isles in the past 2,000 years. The Irish diaspora is exceptional in having well over ten times the source population living abroad. I don't get the idea of L21 marching across Europe like Lemmings and stopping at the Atlantic.

Migrations? Aside from some of the Britons settling in Armorica at the close of the Roman Period, what actual migrations from the British Isles to the Continent took place? The establishment of some Irish monasteries on the Continent hardly constitutes a migration. With regard to the Isles and the European Continent, by far most of the population flow has been into the former from the latter and not the other way around.

I think you are making the classic mistake - which I thought we had overcome a few years ago - of equating the place with the greatest frequency of a y haplogroup with its place of likely origin.



Genographic who have the latest NGS Datasets from the Mayo and Asturias projects indicated that M222 could have originated in Mayo. This is Connachta country. We will not know for sure until the details are published and especially the imminent arrival of a wave of new SNPs. Even more enlightening will be a greatly expanded Phylogenetic Tree.

My understanding is that there were three major expansions of R1b. L2 under U152 associated with the Le Tene and Halstatt Iron Age movement dominating Gaul, Centred on Alesia and Halstatt. DF27 a Bronze Age movement with highest frequency in Iberia and the Atlantic facade. L21 a Bronze Age Atlantic Bell Beaker movement and later M222 with highest frequency in The Isles and Ireland and especially the west of Ireland. I have performed detailed analysis of all the L21 SNPs and they tell a similar story.

I have updated my analysis of L21 and downstream SNPs,
DF21, DF49, C4466, L513, Z253, Z255, L1335, DF41.
http://www.pinterest.com/gerardcorcoran/r1b-l21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df21/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df49/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-c4466/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l513/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z253/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-z255/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-l1335/
http://www.pinterest.com/gerardcorcoran/r1b-l21-df13-df41/

The greatly expanded Phylogenetic tree should also show the sequence of expansion.

If the data shows something different, I will be first to embrace it and adjust my understanding accordingly.

Those are lists based on commercial dna testing, which, as is common knowledge, is heavily heavily skewed to the British Isles. I have been through this before and have posted the numbers from FTDNA's database illustrating this fact. FTDNA's y-dna customer base is drawn by far mostly from the Anglosphere. Thus results from the British Isles far outnumber those from the Continent. Nevertheless, continental L21+ results continue to pop up, and most of them bear little if any connection to the British Isles.

JRW
01-22-2014, 03:57 PM
The French archaeologist and linguist Henri Hubert agreed with your basic premise, but he has the Beaker Folk (whom I am taking the liberty of interpreting as a proxy for L21) departing the North Sea coast for Britain in the Bronze Age and leaving the place nearly deserted. Germanic tribes would begin to settle there in about 700 BC and would succeed in pushing the Celts west of the Rhine around 200 BC or so.

So, in my humble opinion, there wasn't much L21 among the Anglo-Saxons; they were mostly U106 and I-M253. L21 had retreated west of the Rhine into northern Gaul long before the Migration Period. That is why Myres et al and Busby et al found L21 at such low levels in Germany. Still, there is some there, representing descent from the Celts who once lived in western and southern Germany. I disagree with the attempts to chalk the L21 on the Continent up to men from the Isles.

As we collect and analyze results from the upcoming tsunami of SNPs with other information, hopefully we will be able to understand L21's German relationship better. That being said, IMHO, if one believes (as I do) L21 arose in what is present-day France, the most likely pathway eastward would have been along the North Sea Coast, where the people carrying it (who in all probability were not exclusively L21) assimilated/confederated with the indigenous peoples (most likely at least I-M253) in the early Bronze age. As U106 started to flow into the area (there is another thread debating the timing of U106's arrival -- if it were not already present), we can see how Kuhn's Nordwestblock theory of a blended Celtic and Germanic culture in the area makes sense. As wave-after-wave of small subsets of the then-resident hybrid (i.e., dynamically hybrid) population moved to the Isles and perhaps also to present day Norway/Sweeden, the continued (albeit uneven) flow of U106 filled-in the void -- from east to west -- eventually reaching the domination in the area post-Völkerwanderung that we observe today.

Not surprisingly (there are always contrary interpretations on these matters), certain modern-day authors do not provide evidence in their tomes to support Hubert's timing of Celtic departure from the area. Malcom Todd in the Early Germans, cites Celtic artifacts and burial practices found in Jutland, Schleswig and near the Rhine as evidence of Celtic influence in the area through the late Iron Age and Norman Mongan in the Menapia Quest asserts that the Menapia (a Cletic tribe) was well entrenched east of the Rhine in the late Iron Age and Roman Period. However, I think the more supportive evidence comes from the pattern of L21 in Germany, which appears to be concentrated along the Rhine and its tributaries (this observation has been the subject of a number of posts). That pattern is very consistent with L21 (and some subclades, no doubt) at the mouth of the Rhine and migrating up-river, likely starting in the Bronze Age with the spread of L21 eastward along the North Sea Coast, and culminating with the Merovingian settlement of the area in the Migration Period. Franconia, after all, got its name for a reason.

jdean
01-22-2014, 03:57 PM
Looking at Mr Heil's STR data it occurs to me that he may be distantly connected to the Z2961 Fancher/Johnson cluster.

If I were to guess I'd say the Fancher/Johnson group originated in North of England or Scotland but if the cluster were old enough to contain Heil all bets to its source are off.

Goodnow (kit no. 127515 in the French Heritage project) looks to be a reasonable posibility for this group and has an ancestry going back to La Rochelle via Quebec, I emailed him last summer about Z2961 but unfortunately didn't hear back

rms2
01-22-2014, 04:17 PM
As we collect and analyze results from the upcoming tsunami of SNPs with other information, hopefully we will be able to understand L21's German relationship better. That being said, IMHO, if one believes (as I do) L21 arose in what is present-day France, the most likely pathway eastward would have been along the North Sea Coast, where the people carrying it (who in all probability were not exclusively L21) assimilated/confederated with the indigenous peoples (most likely at least I-M253) in the early Bronze age. As U106 started to flow into the area (there is another thread debating the timing of U106's arrival -- if it were not already present), we can see how Kuhn's Nordwestblock theory of a blended Celtic and Germanic culture in the area makes sense. As wave-after-wave of small subsets of the then-resident hybrid (i.e., dynamically hybrid) population moved to the Isles and perhaps also to present day Norway/Sweeden, the continued (albeit uneven) flow of U106 filled-in the void -- from east to west -- eventually reaching the domination in the area post-Völkerwanderung that we observe today.

Kuhn's Nordwestblock theory involves a language neither Celtic nor Germanic, but that is really another topic. Last I heard, few linguists were on board with that theory.

I'm not exactly sure what it is you are driving at. No doubt some L21 remained east of the Rhine, but most of it probably moved west with the advance of the Germans. That is born out by dna testing in the region and the stats from both Myres et al and Busby et al, and by, I believe, the results of the Brabant Project.



Not surprisingly (there are always contrary interpretations on these matters), certain modern-day authors do not provide evidence in their tomes to support Hubert's timing of Celtic departure from the area. Malcom Todd in the Early Germans, cites Celtic artifacts and burial practices found in Jutland, Schleswig and near the Rhine as evidence of Celtic influence in the area through the late Iron Age and Norman Mongan in the Menapia Quest asserts that the Menapia (a Cletic tribe) was well entrenched east of the Rhine in the late Iron Age and Roman Period. However, I think the more supportive evidence comes from the pattern of L21 in Germany, which appears to be concentrated along the Rhine and its tributaries (this observation has been the subject of a number of posts). That pattern is very consistent with L21 (and some subclades, no doubt) at the mouth of the Rhine and migrating up-river, likely starting in the Bronze Age with the spread of L21 eastward along the North Sea Coast, and culminating with the Merovingian settlement of the area in the Migration Period. Franconia, after all, got its name for a reason.

Doubtless there were Celtic peoples who remained in Germania to be absorbed by the Germans. If that wasn't so, there would be no L21 or U152, etc., there today.

IMHO, L21 may have once inhabited the North Sea coast but for the most part either migrated to the British Isles with Rhenish Beaker Folk in the Bronze Age or got chased out by the advancing Germans starting in around 700 BC and culminating around 200 BC.

I'm not sure you are trying to connect the Salian Franks to L21, but if so, I disagree. There is little evidence of much L21 among their descendants, at least those among them who speak Dutch and Flemish, which languages stem from Old Low Franconian.

alan
01-22-2014, 04:35 PM
Generally speaking I think the direction of flow of L21 in prehistory was overwhelmingly from continent to the isles and that the best evidence of out of isles population flow is from the post-Roman period. I think as it stands at present that isles to continent movement of L21 looks likely to be largely a low level historic period thing. The evidence for that can only be strong where the L21 on the continent is of an isles clade that dates to the AD period. When you see isles clades of the AD period on the continent that may well be a trace of out of isles historic period movement. However, as it stands, only a small amount of continental L21 looks likely to be of isles origin.

Noone is denying there must be a small trace of isles to continent L21 but IMO its not very interesting as we know that there was a minor diaspora of monks, wild geese etc from history sources so we learn nothing new from it - some may get a patriotic kick along the 'how the Irish saved civilisation' kind of lines but that was really about cultural impact and genetic impact must have been minimal and is not really the point. Far more interesting IMO is how L21 got into the isles in prehistory. That is much more important as L21 became so dominant among the pre-Roman population of the isles and the details of that remain to be explained. That is fundemantal to the yDNA story of the isles.

alan
01-22-2014, 04:56 PM
I think the dominance of L21 in the isles and the way the within-isles cline of L21 so closely fits the degree of Celtic survival indicates that most of the pre-Germanic inhabitants were L21. That in turn is pretty well a very strong indicator that the main crossings to the isles across the channel on the continent were once a lot higher in L21 than they are today and subsequent dilution has taken place on the continent since the main early migration of L21 to the isles. If that hadnt been the case it would be almost impossible to explain how L21 came to have such a strong showing in the isles. The key is probably that L21 lineages dominated the maritime coasts of France and perhaps even the Low Countries back 4500-4000 years ago regardless of the showing today. The way I look at it is noone arrived in post-Mesolithic times in the isles without having a boat. So, all L21 had to do to have complete dominance of the P312 settlement of the isles was to have dominance of the sea going boats in the channel area. I personally think the sort of L21 dominance we see in Brittany still once applied across the whole continental side of the English Channel. That kind of dominance need not even have extended far inland in order to be of huge importance to the settlement of the isles. This L21 dominance of the sea going traffic in the channel may have continued well into the Bronze Age as indicated by the single cultural 'maritory' suggested to have encapsulated both sides of the English channel in Bronze Age. I personally do not think this was undermined until the late Bronze Age when central European influences commenced with Urnfield i around the Low Countries perhaps bring U152. I still believe U106 was contained east of the Rhine until the Iron Age and later.

rms2
01-22-2014, 07:42 PM
. . .

Far more interesting IMO is how L21 got into the isles in prehistory. That is much more important as L21 became so dominant among the pre-Roman population of the isles and the details of that remain to be explained. That is fundemantal to the yDNA story of the isles.

I agree wholeheartedly with that, but I would add that the periodic attribution of continental L21 to Irish monks and Wild Geese, etc., is harmful. It causes some men to avoid going public with their results or joining any of the appropriate projects, and it spreads misinformation that valuable time, effort and resources must be expended on to dispel.

Heber
01-22-2014, 10:54 PM
Mike,

It appears (Rathna) that Britain's DNA may release a 1,000 anon sample Dataset for analysis.
I would be happy to help with the Ireland L21 segment.

"I am not sure if Jim mentioned this to you - but we have decided to release the anonymised YDNA results of 1000 random male chromo2 customers. This will be in the form of an excell spreadsheet, with one column per customer (BD# removed) and 1 row per SNP. We will also indicate the inferred subtype. I am working on this at the moment but it might be a week or so before it is released."

jdean
01-22-2014, 11:41 PM
Mike,

It appears (Rathna) that Britain's DNA may release a 1,000 anon sample Dataset for analysis.
I would be happy to help with the Ireland L21 segment.

"I am not sure if Jim mentioned this to you - but we have decided to release the anonymised YDNA results of 1000 random male chromo2 customers. This will be in the form of an excell spreadsheet, with one column per customer (BD# removed) and 1 row per SNP. We will also indicate the inferred subtype. I am working on this at the moment but it might be a week or so before it is released."

Brilliant !!!!

Can't wait, thanks for the heads up : )

TigerMW
01-22-2014, 11:45 PM
Mike,

It appears (Rathna) that Britain's DNA may release a 1,000 anon sample Dataset for analysis.
I would be happy to help with the Ireland L21 segment.

"I am not sure if Jim mentioned this to you - but we have decided to release the anonymised YDNA results of 1000 random male chromo2 customers. This will be in the form of an excell spreadsheet, with one column per customer (BD# removed) and 1 row per SNP. We will also indicate the inferred subtype. I am working on this at the moment but it might be a week or so before it is released."

Yes. I was told.

To go the next step, I'd love to be able to associate kit #s or Ysearch IDs with these anonymised samples. Perhaps we can just get people to post this information... their kit #s with the row #/index # or whatever from Jim Wilson's report.

Unfortunately, privacy and other concerns some times get in the way of disclosure, but at least this eliminates the process of people downloading their raw results, when they get them, then uploading them again to the Yahoo group. All one has to do is post their FTDNA or Ysearch ID #.

Heber
01-22-2014, 11:49 PM
I wish Genographic would do the same with their Mayo and Asturias Datasets. We have so much to learn from this new information. Innovation happens when information is liberated.

rossa
01-22-2014, 11:51 PM
I wish Genographic would do the same with their Mayo and Asturias Datasets. We have so much to learn from this new information. Innovation happens when information is liberated.

Does anyone know what their plans are for the Mayo data? Are they just going to sit on it, or at the very least give some more detail on it?

Heber
01-23-2014, 12:09 AM
Does anyone know what their plans are for the Mayo data? Are they just going to sit on it, or at the very least give some more detail on it?

Asturias data was collected last May but no announcements to date.
Mayo data was collected last July and very broad results announced in November.
http://www.pinterest.com/gerardcorcoran/genographic/
This information has a very short shelf life so I don't understand the rationale for delaying publication unless it is the usual peer review cycle or commercial issues re Geno 3.0.

It would be great if they published Asturias and Mayo at the same time as the La Brana publication next week. My latest date for FGC is 1st March. That is about the same time frame as Big Y.

MacEochaidh
01-23-2014, 02:25 AM
One thing to remember about Irish and/or Scottish monasteries is that they were populated by monks who had taken vows of poverty, chastity, and obedience. In other words, the monks were supposed to be celibate. That is not to say they never lapsed and gave in to temptation, but the idea that they had a real, detectable genetic impact on the Continent seems to me to really strain credulity.

I mentioned many times before that these were Monastic Communities. They were centers of religion and education. The communities were populated by families and individuals outside of the religious orders. I think people get the idea of a church, a few buildings and 20 celibate Monks. The Universities alone were magnificent and are some of the first major Universities in European Countries.

ADDED: Plus, some of the Saints connected to Religious Centers had thousands of followers from all over. I remember visiting Ferns with my cousin, Mylie, and being told that St. Aidan (Mogue), the first Bishop of Ferns, was like a "Rock Star" of his age. He was from Cavan and I believe he had studied in Wales. He had followers from all over Ireland and Wales who moved to be in the Diocese of Ferns. Some Irish Saints on the Continent are still big news and I am certain they had followers from the Isles in their time. The name Aidan is still popular in Wexford as was Mogue, and its ridiculous Anglicized form "Moses", a few generations back (my 2x and 3x great grandfathers).

rms2
01-23-2014, 02:36 PM
I mentioned many times before that these were Monastic Communities. They were centers of religion and education. The communities were populated by families and individuals outside of the religious orders. I think people get the idea of a church, a few buildings and 20 celibate Monks. The Universities alone were magnificent and are some of the first major Universities in European Countries.

ADDED: Plus, some of the Saints connected to Religious Centers had thousands of followers from all over. I remember visiting Ferns with my cousin, Mylie, and being told that St. Aidan (Mogue), the first Bishop of Ferns, was like a "Rock Star" of his age. He was from Cavan and I believe he had studied in Wales. He had followers from all over Ireland and Wales who moved to be in the Diocese of Ferns. Some Irish Saints on the Continent are still big news and I am certain they had followers from the Isles in their time. The name Aidan is still popular in Wexford as was Mogue, and its ridiculous Anglicized form "Moses", a few generations back (my 2x and 3x great grandfathers).

You are mixing a couple of things: monastic communities and monks and saints in the Isles and monastic communities and monks and saints on the Continent. Are you maintaining that Irish monastic communities on the Continent were surrounded by large populations of non-celibate people (i.e., people other than monks) from the Isles, enough to make a significant genetic impact?

Soon after the discovery of L21 in late 2008, a number of people came forward with the same basic idea, rooted in the common error that the place with the highest frequency of a y haplogroup must certainly be its place of origin. That idea is that L21 everywhere is to be chalked up to the Irish or, more broadly, to men from the British Isles. No matter how many continental L21s turned up or continue to turn up, no matter how divergent their haplotypes, or the fact that some of them are not even DF13+, no matter that they turn up in a database heavily skewed to the Isles, still that idea persists.

It seems to me too that it is advanced with little in the way of evidence and more as an insinuation, like this: after all, there were Irishmen on the Continent.

Some people innocently advanced the idea that L21 spread to the Continent from Ireland or the British Isles, because they saw the initial results and thought surely L21 must have arisen in the Isles shortly after men returned there from the FC Ice Age Refuge. Others advanced the idea because they wanted there to be a downtrodden, "aboriginal" Isles population to contrast with their own heroic, "invader" ancestors, or to exclude L21 from competition for the title of "true Celts". Some advanced the idea out of a sense of national pride.

But none of them, it seems to me, ever advanced the idea with much evidence and definitely none that is very compelling. On the other hand, we know that 1) the British Isles have been on the receiving end of numerous substantial population movements from the Continent throughout prehistory and history; 2) that very early on Celtic languages were spoken there that certainly did not have their ultimate origin in the Isles; 3) that L21 is found on the Continent, in some places at pretty high frequency and in many cases not at all resembling that found in the Isles; 4) that P312, the immediate ancestor of L21, and the offspring of P312, the close siblings of L21, certainly did not originate in the British Isles; and 5) that Ireland is, as far as we know, nearly 100% DF13+, whereas continental Europe has its share of DF63.

MacEochaidh
01-23-2014, 03:19 PM
You are mixing a couple of things: monastic communities and monks and saints in the Isles and monastic communities and monks and saints on the Continent. Are you maintaining that Irish monastic communities on the Continent were surrounded by large populations of non-celibate people (i.e., people other than monks) from the Isles, enough to make a significant genetic impact?

Soon after the discovery of L21 in late 2008, a number of people came forward with the same basic idea, rooted in the common error that the place with the highest frequency of a y haplogroup must certainly be its place of origin. That idea is that L21 everywhere is to be chalked up to the Irish or, more broadly, to men from the British Isles. No matter how many continental L21s turned up or continue to turn up, no matter how divergent their haplotypes, or the fact that some of them are not even DF13+, no matter that they turn up in a database heavily skewed to the Isles, still that idea persists.

It seems to me too that it is advanced with little in the way of evidence and more as an insinuation, like this: after all, there were Irishmen on the Continent.

Some people advanced the idea that L21 spread to the Continent from Ireland or the British Isles innocently, because they saw the initial results and thought surely L21 must have arisen in the Isles shortly after men returned there from the FC Ice Age Refuge. Others advanced the idea because they wanted there to be a downtrodden, "aboriginal" Isles population to contrast with their own, heroic, "invader" ancestors, or to exclude L21 from competition for the title of "true Celts". Some advanced the idea out of a sense of national pride.

But none of them, it seems to me, ever advanced the idea with much evidence and definitely none that is very compelling. On the other hand, we know that 1) the British Isles have been on the receiving end of numerous substantial population movements from the Continent throughout prehistory and history; 2) that very early on Celtic languages were spoken there that certainly did not have their ultimate origin in the Isles; 3) that L21 is found on the Continent, in some places at pretty high frequency and in many cases not at all resembling that found in the Isles; 4) that P312, the immediate ancestor of L21, and the offspring of P312, the close siblings of L21, certainly did not originate in the British Isles; and 5) that Ireland is, as far as we know, nearly 100% DF13+, whereas continental Europe has its share of DF63.

What I am saying is that some (SOME- meaning "some"/not all/not the bulk of.../not implying origin) L21 and Subclades, including DF49, DF23, and M222 may be attributed to Religious Movements to the Continent. In later years, Religious Movements, such as the Huguenots, have certainly accounted for some (SOME) genetic movement in Europe. Actually, I have heard peopel use the Huguenots as an example of Continental genetic movement into the Isles. Therefore, using the same logic, I believe some (SOME- not all/not the bulk of...'not implying origin) of the early Religious Movements out of the Isles to the Continent may have contributed to some (SOME) genetic test results found today.

The word "some" is often confusing, but a quick look in a dictionary will show it is not a synonym for "origin" or "all". Also, it may be time to stae clearly one more time that I, Miles Douglas Kehoe, do NOT believe that L21 originated in The Isles. I, Miles Douglas Kehoe believe that DF23 originated in the northwest area of what is now France, or at least, came to The Isles from that area.

Watch now as this is turned into my making an argument for an Isles L21 origin. It happens EVERY time. not just SOME of the time :)

JRW
01-23-2014, 03:50 PM
Kuhn's Nordwestblock theory involves a language neither Celtic nor Germanic, but that is really another topic. Last I heard, few linguists were on board with that theory.

I'm not exactly sure what it is you are driving at. No doubt some L21 remained east of the Rhine, but most of it probably moved west with the advance of the Germans. That is born out by dna testing in the region and the stats from both Myres et al and Busby et al, and by, I believe, the results of the Brabant Project.



Doubtless there were Celtic peoples who remained in Germania to be absorbed by the Germans. If that wasn't so, there would be no L21 or U152, etc., there today.

IMHO, L21 may have once inhabited the North Sea coast but for the most part either migrated to the British Isles with Rhenish Beaker Folk in the Bronze Age or got chased out by the advancing Germans starting in around 700 BC and culminating around 200 BC.

I'm not sure you are trying to connect the Salian Franks to L21, but if so, I disagree. There is little evidence of much L21 among their descendants, at least those among them who speak Dutch and Flemish, which languages stem from Old Low Franconian.


All of my comments related to Mike's initial post (and some of the subsequent posts) in this thread. But obviously, I didn't do a very good job of tying my points with Mike's observations. So, here's another try:

1) Martin Heil, although a resident of Hungary, is most likely of German ancestry. There were many (i.e., millions) ethnic Germans residing in eastern Europe in the 19th century. His line most likely moved there in recent history. I gave an example of my own family, where by the mid-19th century, one 18th century L21/DF13 male German from Göttingen had descendants residing in present-day Poland, Slovakia, and Hungary (not to mention the UK, Canada, and the US).

2) There ARE descendants from the Isles in what is present day Germany (and present day Poland for that matter). In fact, there are not just articles, but books written on the subject. However, by reviewing the surnames and/or genetic distances of the "German" with Isles population, I believe the vast majority of these can be readily identified (and Heil does not appear to be one). Descendants of Isles immigrants constitute a very small proportion of L21 Germans.

2) The story of L21 (and a yet unknown set of its subclades) in what is present-day Germany is a complex one. The best explanation, IMHO, given what in known today from synthesizing information from multiple-disciplines, is that L21 originated in what is present-day France (most likely northern), and spread from there in multiple directions -- including across The Channel and along the North Sea Coast to Jutland (I believe they had strong sea-faring skills). This happened in the early Bronze Age.

3) Along the North Sea Coast, L21 integrated with the indigenous population, the composition of which is a topic of debate (e.g., U106). The point here, as it relates to the Isles, is that some L21 found its way to the Isles through this route. So, one (hypothetically) might find a L21 Dane in the same (very old) subclade as say, someone from Kent.

4) The establishment of L21 along the North Sea Coast positioned it to move up the Rhine and its tributaries, which it did, IMHO, starting in the Bronze Age. This is where my view differs from the consensus view that L21 travelled down the Rhine from the Rhone or Danube basins, or wherever. The general distribution pattern of L21 in Germany is along the Rhine and its tributaries. And, my point about the Franks, is that the movement of peoples (who have some L21 among them) up the Rhine continued through the Iron Age into the historic period. I would assert that the proportion of L21 along the coast changed over time as the flow of U106 into the area gained dominance, so it's anyone's guess as to the haplogroup composition of the Franks, Angles, Saxons, etc.. The one point that I would argue is that none was composed of a single haplogroup.

5) There were several posts mentioning that Heil's ancestors might be from Alsace-Lorraine (Moselle River basin) or he might be related to a Heil from Frankfurt (on the Main River). Either way, he appears to be an example of someone who had a L21 ancestor(s) (or subclade thereof), who moved up the Rhine (and then the Moselle or Main) from the North Sea Coast. His relationship to his Isles kin is that they shared a common ancestor along the North Sea Coast. As we get more SNP data, I believe this story will be repeated again and again for German L21.

6) A point that I didn't make, but feel compelled to do so for the benefit of those new to this area of study (and I know you know better), is that the snap-shot of an area's haplogroup composition today is information that needs to be interpreted and placed into context with other information, starting with written history of the area. It is not a conclusion. The haplogroup composition of the Netherlands and Belgium (particularly the sub-areas thereof) today is informative, but not in anyway determinative of what it was 2000 years ago.

Dubhthach
01-23-2014, 03:59 PM
Perhaps it would be nice if we could get back to talking about M222 and it's origin. Personally I wouldn't be surprised if DF23 arose on the contient (and thus DF49 before it). The question really is where did Z2961 arise and then it's offspring M222. My bet is both probably arose in what is now Northern Britain which was "Insular Celtic" linguistically at the time of some suggested dates of age etc.

Chromo2 is showing a very early branching point in M222 with snp called S7073, this is the DF13 of M222. So far there is only one anonymous M222+/S7073- whose most recent ancestral origin is in Fife Scotland. Within S7073 there is a further differentiation regarding those who are S658+ and those who are S658-. Several of Scottish M222+ are confirmed S7073+/S658-, whereas there are alot of Irish surnames within S658+

Some of these SNP's on Chromo2 under M222 definitely arose within Ireland.

Heil needs to be tested for Z2961, he'll probably be negative but it would at least seperate him into DF23* as oppose to DF23+/Z2961+/M222- block.

-Paul
(DF41+)

jdean
01-23-2014, 04:20 PM
Heil needs to be tested for Z2961, he'll probably be negative but it would at least seperate him into DF23* as oppose to DF23+/Z2961+/M222- block.

(DF41+)

He ordered it last night, anybody fancy setting up a book on the result : )

Personal I think there's quite a good chance he'll be positive.

BTW (and unrelated to the above comment) we already have a French Z2961 via Monsieur Leprost kit no. N92711

Eochaidh
01-24-2014, 02:18 AM
Several of Scottish M222+ are confirmed S7073+/S658-, whereas there are alot of Irish surnames within S658+
There is one Irish surname, from Louth, who is S7073+/S658-. That is me. All six of us are FGC4077/FGC4078+.

TigerMW
01-24-2014, 02:42 AM
There is one Irish surname, from Louth, who is S7073+/S658-. That is me. All six of us are FGC4077/FGC4078+.
What are the kit #s for the FGC4077/FGC4078+ people. I'd like to make sure that is properly reflected in the spreadsheet. I know Kennedy does not have STRs but what about the other five?

David Wilson
01-24-2014, 12:23 PM
The following are S568-, S658- and FGC4077/FGC4078+

WILSON 8999
Lominac-SUDDUTH N14949
REID 205699
HOY N112505

The fifth individual assigned to that group is represented by a PGP genotype but his name is not known.

LAMONT 108794 is in his own bracket S568 below S7073. Kennedy's tree excludes him from membership in FGC4077/4078, but sitting here in the California dark before coffee I don't recall if those those markers were actually tested for him.

Dubhthach
01-24-2014, 12:44 PM
David,

Thanks for that both Hoy and Lominac-SUDDUTH are in the Ireland project. I've created a subgroup that contains both of them:
"Group16-1h5a R1b-FGC4077/FGC4078+ (M222+) -- Chromo2 confirmed"

-Paul
(DF41+)

jdean
03-29-2014, 06:24 PM
Looking at Mr Heil's STR data it occurs to me that he may be distantly connected to the Z2961 Fancher/Johnson cluster.

If I were to guess I'd say the Fancher/Johnson group originated in North of England or Scotland but if the cluster were old enough to contain Heil all bets to its source are off.

Goodnow (kit no. 127515 in the French Heritage project) looks to be a reasonable posibility for this group and has an ancestry going back to La Rochelle via Quebec, I emailed him last summer about Z2961 but unfortunately didn't hear back

Mr Heil lost his DF23* status last night : )

As previously noted he only has one real match is a Hiles with ancestry from Frankfurt.

Think I'm having to go for third time lucky with Mr Goodnow, a BigY from Mr Heil would be rather nice as well : )

Rory Cain
09-03-2015, 08:25 PM
Perhaps it would be nice if we could get back to talking about M222 and it's origin. Personally I wouldn't be surprised if DF23 arose on the contient (and thus DF49 before it). The question really is where did Z2961 arise and then it's offspring M222. My bet is both probably arose in what is now Northern Britain which was "Insular Celtic" linguistically at the time of some suggested dates of age etc...

-Paul
(DF41+)

Would I be off-topic if I speculated that M222 might show some correlation with the Selgovae and/or Brigantes of the Anglo-Scots Border; L513 with the Novantes territory in Galloway & DF21 with Damnonii territory in Strathclyde and Lomondside?

These three SNPs are represented in Ireland too. There has been debate about the direction if travel. M222 and DF21 may well have gone from Scotland to Ireland, the timing perhaps coinciding with Roman Legions pressing them. L513 looks older with it's distribution up the west Coast from Kerry to the Scottish Isles, some of which movement may perhaps have been bump-on effect out of Ireland from the arrival if M222 and DF21?

Just an attempt to put events into context. Maybe it's still to early to do do.

Dubhthach
09-07-2015, 08:40 AM
The issue we have at moment is that ScotlandsDNA/IrelandsDNA hasn't release any data with regards to division of various M222 subclades within their testing database. So we don't know for example if the incidence of M222 in their database that is from Scottish men belong to upstream or parallel clades to incidence seen in Ireland.

They did release a generic map, but it seems to be just all of M222, plus caveat that we have no idea what the sample size is per region, 30% of 20 for example could be sample bias compare to say 30% of 300.

http://compsoc.nuigalway.ie/~dubhthach/DNA/m222_spread.png

The current M222 tree looks somewhat like this:
http://www.kennedydna.com/M222_tree.png

Dubhthach
09-07-2015, 08:43 AM
There's something like 23 SNP's in Chromo2 test that are equivalent to M222, obviously the line underwent a severe bottleneck for that many SNP's to accumulate in one lineage before expanding:

https://dl.dropboxusercontent.com/u/10258680/Draft_DF49xM222_Tree_v20.png

miiser
09-07-2015, 08:52 AM
There's something like 23 SNP's in Chromo2 test that are equivalent to M222, obviously the line underwent a severe bottleneck for that many SNP's to accumulate in one lineage before expanding:


Or the lineage experienced 23 SNP mutations in a brief time period. There's no data based evidence to show that a cluster of SNPs didn't or couldn't occur within a small number of generations.

Dubhthach
09-07-2015, 09:18 AM
Or the lineage experienced 23 SNP mutations in a brief time period. There's no data based evidence to show that a cluster of SNPs didn't or couldn't occur within a small number of generations.

Sure well if you had 2-3 per generation it would only take 10 generations to get to that level (250 years)

Well one of things we need a study of is a large group of say 4-5th cousins (20+) who have their Y Sequenced which will allow us to generate a more accurate figure on SNP mutation rates. Personally I'm not too sure about some of averages that people have been throwing about when it comes to mutation rates for Y-SNP's. (eg. average of 1 per 100 years etc.)

What's evident though with M222 is after this point where you have this block (which if you look at BigY/FGC testing is actually 42 snp's!) that you see rapid expansion with many of SNP's only show either no equivalents or at most 2 at any given level.

Prime example is under DF105 (which is equivalent to DF109), we see:
S588 (no equivalents)
DF85 (2 equivalents)
A259 (no equivalents)
FGC4133 (2 equivalents)
A1206 (no equivalents)
A223 (one equivalent)

miiser
09-07-2015, 10:28 AM
I agree, a study of closely related people would be informative. If SNPs typically occur as single, isolated events, then we should expect the SNPs to be evenly distributed throughout all the lineages. But if SNP runs occur at a significant frequency, then we should see a larger number of lineages having no observed SNPs and some minority of lineages with excessive SNPs.

I'm not so much concerned that the average rate being used is wrong. I think it's close to correct, at least for a limited time range. The concern is that assuming SNPs are singular, isolated events leads to incorrect conclusions regarding the timing of specific branches and the growth rate of haplogroups.

For a long time, most people have been assuming that SNPs occur at a regular frequency as single events, and consequently interpreting phylogenetically equivalent SNP blocks as population bottlenecks and multibranch nodes as growth bursts. But this is wrong thinking. These observations can more elegantly be explained by mutation runs.

The test of any hypothesis is whether it fits the data. It is possible to evaluate, using the data we already have, whether the model of singular isolated SNPs or the model of occasional SNP runs better fits the data.

And, in fact, there is already significant evidence in the data to clearly demonstrate that SNPs do tend to be clustered together in time. Firstly, there is the number of branches per node seen throughout the tree. If the SNPs are randomly distributed as single events, we should expect the typical node of the tree to have 2 branches, with occasional nodes having a greater number of branches. Any simple population model shows this to be the case. But the tree that has developed in L21, for example, has an average node size of more like 4 branches. This is not consistent with singular SNP mutations, but is consistent with SNP runs. The SNPs being clustered together effectively collapse what would otherwise be multiple nodes of 2 branches each into a single node of multiple branches. The ~4 branches per node seen in the tree is consistent with an average run size of around 3 or 4 SNPs per run.

If one assumes that SNPs occur as singular isolated events, then one is forced to conclude that the L21 tree has grown almost EXCLUSIVELY by a small number of rapid spurts, separated by long intervals of no growth. This is an awkward and unnatural interpretation of the data. A much more natural interpretation, more consistent with the archaeological and historical data, is that growth has been mostly smooth, but the SNPs have occurred at an uneven rate.

Secondly, we have the large variation of the number of SNPs in parallel branches. The branch-to-branch variation observed in the tree far exceeds what one should expect from randomly distributed single SNPs. You have numerous cases where one branch has, let's say, 15 SNPs, and its sibling branch has perhaps only 5. This large variation is not consistent with SNPs occurring as single events, but is consistent with SNPs being clustered together. The clustering has the effect of increasing the variation in the number of SNPs from one branch to another.

So the data we already have is consistent with the hypothesis of SNPs sometimes occurring as runs, and contradicts the hypothesis of SNPs occurring mainly as singular, isolated events.

Finally, there are a number of known mechanisms that can readily explain how or why SNPs may occur as runs. The age of the father is proven to increase the mutation rate. This alone could easily generate a sizable run within a few generations of long lived fathers. There are a variety of effects that can cause DNA damage and dramatically increase the mutation rate for a limited duration or within a particular culture or locale: cosmic gamma ray bursts, toxin exposure (such as arsenic from mining/metallurgy), various disease states that cause oxidative stress resulting in DNA damage (vitamin C deficiency, etc.). Finally, there are the mechanical effects of the DNA molecule itself - an initial mutation can cause a structural instability, resulting in multiple subsequent mutations until a new stable configuration is achieved. There are some SNPs in the data which are spatially near to each other on the chromosome, indicating that there is in fact a causal relationship between mutations in at least some cases.

With all of these possible effects combined, I would expect the number of SNPs per run to be something like a logarithmic distribution. The mode of the run length would probably still be one. But there would be occasional runs of significant length, giving an average run length of maybe three or four. If this is the case then, using the accepted average rate, a single SNP or SNP run would occur once every ~500 years or so, rather than once every ~149 years as is currently assumed by most people. And this would generate exactly the sort of tree structure that we observe. 23 or 42 SNP mutations very well could have occurred in M222 within just a few generations.

I agree that a robust study of closely related individuals would be beneficial. But there is already overwhelming evidence against SNPs being evenly distributed as single, isolated events. People need to move past the assumption that phylogenetically equivalent blocks of SNPs imply bottlenecks and multibranch nodes imply growth spurts. This may be true in some cases, but it is not generally true.

MacUalraig
09-07-2015, 12:16 PM
I think your point is a good one which is worth considering, although maybe DF49 diversity is sufficiently high to make the numbers fit...?

razyn
09-07-2015, 12:56 PM
So the data we already have is consistent with the hypothesis of SNPs sometimes occurring as runs, and contradicts the hypothesis of SNPs occurring mainly as singular, isolated events.
I have previously agreed with miiser (or vice versa) about this; just want to mention here that the issue is much more widespread than one specific branch under L21. And the main problem with this community's ignoring it pretty consistently, over time and across the globe, is that (if true -- and the better data we get, the more true it appears) it contradicts an underlying assumption of the technique of estimating age by counting SNPs.

Estimating it by counting the forks that we know led to isolable branches would not have that drawback. But that is barely being done by anybody. Seems to me, Marko Heinila was doing that, before he dropped off the radar about three years ago. (Sorry if I have misspelled either of his names.) We have identified a lot more branching, by now.

If I have any disagreement with miiser it is a little doubt that these multiple "equivalents" happened in runs, i.e. several at a time for several generations. I suspect that a level or branch with many equivalent SNPs is evidence of an event of some sort, that happened to one man; hence for practical purposes, at one time and in one place. (The latter two modifying phrases aren't necessary, just probable.) All of the people who still bear all of those mutations (in distinction from fellow descendants of their next larger haplogroup) descend from that guy -- who wasn't sterilized by the big event, and became that line's Abraham.

Here is an earlier instance of the discussion, on a more broadly based forum (though it's still just YDNA in the R haplogroup -- and the caveat applies to any such effort to plot the chronology of phylogenetic trees): http://www.anthrogenica.com/showthread.php?4255-Genetic-Drift-Sharing-Bell-Beaker-amp-WHG&p=78296&viewfull=1#post78296

Jean M
09-07-2015, 12:58 PM
Over on that thread about Jean's book, you mentioned her idea that M222 could represent the arrival in the Isles of some continental settlers bearing La Tene culture and skills.

This was not my idea. (I missed this at the time it was posted.) Mike has it right in post http://www.anthrogenica.com/showthread.php?1907-Origins-of-M222-DF23-the-NW-Irish-and-connections-to-Niall-of-the-Nine-Hostages&p=27895&viewfull=1#post27895



I think the general idea is that M222 came into Ireland from Northern Britain with La Tène.

I could be wrong of course.

miiser
09-07-2015, 02:02 PM
If I have any disagreement with miiser it is a little doubt that these multiple "equivalents" happened in runs, i.e. several at a time for several generations. I suspect that a level or branch with many equivalent SNPs is evidence of an event of some sort, that happened to one man; hence for practical purposes, at one time and in one place. (The latter two modifying phrases aren't necessary, just probable.) All of the people who still bear all of those mutations (in distinction from fellow descendants of their next larger haplogroup) descend from that guy -- who wasn't sterilized by the big event, and became that line's Abraham.

I think this is just a differing understanding of terminology. When I say "run", my intent was to include multiple SNPs occurring in a single individual as a subset of such runs. I don't strongly favor single generation events over multi generational runs, as you do. But I don't intend to exclude them either.

I'm focused mainly on the data analysis, and I think the data is consistent with either scenario. The run just has to occur within a small enough number of generations (1 or a few) for the equivalent block to not be broken by a new surviving branch. Or if it is broken, it will appear as 2 separate blocks of equivalents.

Some types of effects, such as a GRB, occur in a single generation by definition, and should have a global influence across all branches. But some effects, such as molecular instability, are dependent on the number of cellular division generations, requiring a certain number of step-wise mutations in order for the "kink" to work itself out. In those cases, the line between father and son is an arbitrary division in terms of opportunities for mutations to occur. The instability might resolve within a single individual, or it might extend into the next generation. Cultural effects, such as a dietary deficiency or toxic exposure, might conceivably affect everyone within a particular locale for multiple generations. And an elderly father may be just a one off occurrence, or a broader trend associated with prosperity or cultural factors.

I suppose if one were to thoroughly study equivalent blocks in multiple branches, comparing time and location, one might discover patterns that suggest some of them are associated with a specific time, event, locale, cultural practice, etc. I don't have the expertise or data to say which of these effects are most significant or to guess whether or not most runs would be single generation. I know you've mentioned mining and arsenic before as an example of a known practice possibly correlating with an increased mutation rate, and this strikes me as reasonably likely.


And the main problem with this community's ignoring it pretty consistently, over time and across the globe, is that (if true -- and the better data we get, the more true it appears) it contradicts an underlying assumption of the technique of estimating age by counting SNPs.

I am also irked by the persistent ignoring. And I am especially irked by the people who will undoubtedly respond, "No, you don't understand. Statistics! Distributions! Variance! Big words! Averaging! Calibration! Law of large numbers! More big words!!! How dare you challenge my authority! I know about SNP runs. In fact I first noted this phenomenon 2 years ago!" . . . and then a week later they will make another post about equivalent blocks and thousand year bottlenecks. My goal in writing this final paragraph is not to insult anyone, but to pre-emptively shame anyone who would otherwise have responded such, and to make it embarrassingly difficult for anyone to recycle the same old response and continue incorrectly utilizing SNP counting by ignoring the uneven distribution of SNPs.

Dubhthach
09-07-2015, 02:13 PM
If we take most of the TMRCA calculations with a pinch of salt, expansion of major M222 lineages appear to occur in period 1600-1800 years ago. There is an interesting correlation here with what appears to have happened in Irish history during this period. With massive expansion of specific genealogical lineages appearing to be period 400-700AD.

Given nature of Irish society in medieval period, it was very easy for lineage to undergo rapid expansion. Here's an extract form Nichols "Gaelic and Gaelicized Ireland in the Middle ages" (by far the best introduction to Gaelic Ireland)



One of the most important phenomena in a clan-based society is that of
expansion from the top downwards. The seventeenth-century Irish scholar and
genealogist Dualtagh Mac Firbisigh remarked that 'as the sons and families of
the rulers multiplied, so their subjects and followers were squeezed out and
withered away; and this phenomenon, the expansion of the ruling or dominant
stocks at the expense of the remainder, is a normal feature in societies of this
type. It has been observed of the modern Basotho of South Africa that 'there is
a constant displacement of commoners by royals [i.e. members of the royal clan]
and of collateral royals by the direct descendants of the ruling prince;, and
this could have been said without adaptation , of any important Gaelic or
Gaelicized lordship of late medieval Ireland.

In Fermanagh, for example the kingship of the Maguires began only with the
accession of Donn Mór in 1282 and the ramification of the family - with the
exception of one or two small and territorially unimportant septs - began with
the sons of the same man. the spread of his descendants can be seen by the
genealogical tract called Geinelaighe Fhearmanach; by 1607 they must have been
in the possesion of at least three-quarters of the total soil of Fermanagh,
having displaced or reduced the clans which had previously held it. The rate
which an Irish clan could itself must not be underestimated. Tulrlough an fhíona
O'Donnell, lord of Tirconnell (d. 1423) had eighteen sons (by ten different
women) and fifty-nine grandsons in the male line. Mulmora O'Reilly, the lord of
East Brefny, who died in 1566, had at least fifty-eight O'Reilly grandsons.
Philip Maguire, lord of Fermanagh (d. 1395) had twenty sons by eight mothers,
and we know of at least fifty grandsons. Oliver Burke of Tirawley (two of whose
became Lower Mac William although he himself had never held that position) left
at least thirty-eight grandsons in the male line. Irish law drew no distinction
in matters of inheritance between the legitimate and the illegitimate and
permitted the affiliation of children by their mother's declaration (see Chapter
4), and the general sexual permissiveness of medieval Irish society must have
allowed a rate of multiplication approaching that which is permitted by the
polygyny practised in, for instance, the clan societies of southern Africa
already cited.


Now with regards to major subclades of DF105 under M222, we are seeing some sort of geographic/"historic genealogical" expansion going on of various clades. In Irish case majority of men tested on stuff like M222 bundle are showing up DF105+ (or positive for major branch such as DF85, S588, A259/A260 etc.), though some for parallel branches as well.

Of course Irish history was subject to mass-scale rewriting during the 7th-8th century. TM Charles Edwards for example talks about the redacting of the careers of Cairbre mac Néill and Fiacha mac Néill, mainly due to fact that within the Southern Uí Néill overlordship that the power had switched to Cland Cholmáin and thus "dynasts" of Cenél Cairpre and Cenél Fiacha had been sidelined from "province overlodship"

http://compsoc.nuigalway.ie/~dubhthach/DNA/eci_ui_neill.html

Just to tie in with mention about "Older Fathers" a good example from history is Tairrdelbach Ua Conchobair (King of Connacht -- later High-King of Ireland). His youngest son Cathal Crobhdearg Ua Conchobair was born when the bould Tairrdelbach was 65. Cathal lived to age of 71 and the current "O'Conor Don" is direct line descendant of Cathal.

miiser
09-07-2015, 02:46 PM
I think your point is a good one which is worth considering, although maybe DF49 diversity is sufficiently high to make the numbers fit...?

The tree structure of DF49 xM222 is more like one would expect from randomly distributed, single SNPs. But it still does include some large equivalent blocks that are likely at least partially attributable to runs. Growth rate could be a factor for the typically smaller node size if DF49 xM222 was less successful than M222. I'm not arguing that NONE of L21's multibranch nodes are due to very prolific families. Growth spurts could possibly explain a PORTION of the starburst structure in L21, just not all of it.

George Chandler
09-07-2015, 02:49 PM
I agree, a study of closely related people would be informative. If SNPs typically occur as single, isolated events, then we should expect the SNPs to be evenly distributed throughout all the lineages. But if SNP runs occur at a significant frequency, then we should see a larger number of lineages having no observed SNPs and some minority of lineages with excessive SNPs.

I'm not so much concerned that the average rate being used is wrong. I think it's close to correct, at least for a limited time range. The concern is that assuming SNPs are singular, isolated events leads to incorrect conclusions regarding the timing of specific branches and the growth rate of haplogroups.

For a long time, most people have been assuming that SNPs occur at a regular frequency as single events, and consequently interpreting phylogenetically equivalent SNP blocks as population bottlenecks and multibranch nodes as growth bursts. But this is wrong thinking. These observations can more elegantly be explained by mutation runs.

The test of any hypothesis is whether it fits the data. It is possible to evaluate, using the data we already have, whether the model of singular isolated SNPs or the model of occasional SNP runs better fits the data.

And, in fact, there is already significant evidence in the data to clearly demonstrate that SNPs do tend to be clustered together in time. Firstly, there is the number of branches per node seen throughout the tree. If the SNPs are randomly distributed as single events, we should expect the typical node of the tree to have 2 branches, with occasional nodes having a greater number of branches. Any simple population model shows this to be the case. But the tree that has developed in L21, for example, has an average node size of more like 4 branches. This is not consistent with singular SNP mutations, but is consistent with SNP runs. The SNPs being clustered together effectively collapse what would otherwise be multiple nodes of 2 branches each into a single node of multiple branches. The ~4 branches per node seen in the tree is consistent with an average run size of around 3 or 4 SNPs per run.

If one assumes that SNPs occur as singular isolated events, then one is forced to conclude that the L21 tree has grown almost EXCLUSIVELY by a small number of rapid spurts, separated by long intervals of no growth. This is an awkward and unnatural interpretation of the data. A much more natural interpretation, more consistent with the archaeological and historical data, is that growth has been mostly smooth, but the SNPs have occurred at an uneven rate.

Secondly, we have the large variation of the number of SNPs in parallel branches. The branch-to-branch variation observed in the tree far exceeds what one should expect from randomly distributed single SNPs. You have numerous cases where one branch has, let's say, 15 SNPs, and its sibling branch has perhaps only 5. This large variation is not consistent with SNPs occurring as single events, but is consistent with SNPs being clustered together. The clustering has the effect of increasing the variation in the number of SNPs from one branch to another.

So the data we already have is consistent with the hypothesis of SNPs sometimes occurring as runs, and contradicts the hypothesis of SNPs occurring mainly as singular, isolated events.

Finally, there are a number of known mechanisms that can readily explain how or why SNPs may occur as runs. The age of the father is proven to increase the mutation rate. This alone could easily generate a sizable run within a few generations of long lived fathers. There are a variety of effects that can cause DNA damage and dramatically increase the mutation rate for a limited duration or within a particular culture or locale: cosmic gamma ray bursts, toxin exposure (such as arsenic from mining/metallurgy), various disease states that cause oxidative stress resulting in DNA damage (vitamin C deficiency, etc.). Finally, there are the mechanical effects of the DNA molecule itself - an initial mutation can cause a structural instability, resulting in multiple subsequent mutations until a new stable configuration is achieved. There are some SNPs in the data which are spatially near to each other on the chromosome, indicating that there is in fact a causal relationship between mutations in at least some cases.

With all of these possible effects combined, I would expect the number of SNPs per run to be something like a logarithmic distribution. The mode of the run length would probably still be one. But there would be occasional runs of significant length, giving an average run length of maybe three or four. If this is the case then, using the accepted average rate, a single SNP or SNP run would occur once every ~500 years or so, rather than once every ~149 years as is currently assumed by most people. And this would generate exactly the sort of tree structure that we observe. 23 or 42 SNP mutations very well could have occurred in M222 within just a few generations.

I agree that a robust study of closely related individuals would be beneficial. But there is already overwhelming evidence against SNPs being evenly distributed as single, isolated events. People need to move past the assumption that phylogenetically equivalent blocks of SNPs imply bottlenecks and multibranch nodes imply growth spurts. This may be true in some cases, but it is not generally true.


Much of what you're suggesting depends on the test type and then how the SNP's are analyzed and vetted. A person can toss up all of the SNP's from their Full Genomes or Big Y results and display them which is one way of doing it, but if you're displaying SNP positions from problematic locations you're going to get a different average for the number of years per SNP than if you remove the problematic ones. If you remove the problematic ones you will see gaps and runs (if I'm understanding your point correctly). You still can see gaps and runs of SNP's in the data before removing the problematic ones as well but it's less obvious unless you're researching older proven lines using genealogy. I'm still getting the average of about 139 years between SNP but some are close to 200 year (or likely more) and some are under 100 years. You're right that there can be 500 years between two SNP's but it's likely not because there hasn't been a mutation it's because there are several which have been dropped because of reliability issues. I would be interested if you're ever able to capture multiple Y SNP mutations in a single person's sample which are reliable. If you do find that let me know.

In terms of genetic bottlenecks..history is full of events such as plagues, wars and geologic disasters. Saying the data shows a population bottleneck will again depend on the person who analyzes the data and how.

George

miiser
09-07-2015, 03:00 PM
If we take most of the TMRCA calculations with a pinch of salt, expansion of major M222 lineages appear to occur in period 1600-1800 years ago. There is an interesting correlation here with what appears to have happened in Irish history during this period. With massive expansion of specific genealogical lineages appearing to be period 400-700AD.

Given nature of Irish society in medieval period, it was very easy for lineage to undergo rapid expansion.

I don't deny that rapid growth can occur. But assuming only randomly distributed single SNPs requires you to not only believe that a growth spurt occurred. It requires you to believe that nearly all the growth occurred within brief spurts of just a few generations - so that there was no time for any SNPs to break a multibranch node - and then there was NO significant growth for a long time, allowing a block of equivalent to SNPs to form, and then another brief spurt, and another drought, etc. Assuming evenly distributed single SNPs, the only compatible interpretation is that nearly all the growth occurred through just a few brief spurts, with nothing in between.

It defies belief. And there is no reason to embrace such a difficult position when there is already such ample evidence that the SNPs are NOT evenly distributed.

miiser
09-07-2015, 03:13 PM
Much of what you're suggesting depends on the test type and then how the SNP's are analyzed and vetted. A person can toss up all of the SNP's from their Full Genomes or Big Y results and display them which is one way of doing it, but if you're displaying SNP positions from problematic locations you're going to get a different average for the number of years per SNP than if you remove the problematic ones. If you remove the problematic ones you will see gaps and runs (if I'm understanding your point correctly). You still can see gaps and runs of SNP's in the data before removing the problematic ones as well but it's less obvious unless you're researching older proven lines using genealogy. I'm still getting the average of about 139 years between SNP but some are close to 200 year (or likely more) and some are under 100 years. You're right that there can be 500 years between two SNP's but it's likely not because there hasn't been a mutation it's because there are several which have been dropped because of reliability issues. I would be interested if you're ever able to capture multiple Y SNP mutations in a single person's sample which are reliable. If you do find that let me know.

In terms of genetic bottlenecks..history is full of events such as plagues, wars and geologic disasters. Saying the data shows a population bottleneck will again depend on the person who analyzes the data and how.

George

I understand that SNP test methods vary, and this is why I use the term "observed" mutations when I discuss such analysis. My analysis does not assume that every SNP is detected. It assumes that a Big Y test method is used, which is true for nearly all the available test samples. There will be some variation in the number of base pairs sequenced and the coverage from one test to another. This difference is not able to account for the varying number of SNPs seen from one branch to another. The difference of SNPs from one branch to another is a difference of SNPs observed per BP sequenced. The method used is standardized, and, on average, the fraction of detected SNPs should be similar from one sample to another throughout the tree in different branches. No matter how you do the analysis, if you do an apples to apples comparison, there is excessive variation in the count beyond what is expected for single, random mutations, beyond what can be accounted for by differences in test method.

The test method and missed SNPs is wholly unable to account for large blocks of equivalent SNPs. Test method is not a factor here.

You ask whether I am "able to capture multiple Y SNP mutations in a single person's sample which are reliable". There are numerous people who have multiple novel, private SNPs with good confidence from Big Y, if that is what you are asking for. If you are asking for a measure of the number of SNPs between father and son, the answer is that we don't have the data because very few people find a reason to purchase a NGS test for both father and son.

I agree that there have been bottlenecks. I also believe that bottlenecks and spurts are the exception, not the norm. Believing that SNPs occur as single, randomly distributed events does not just require you to believe that there have been bottlenecks. It requires you to believe that nearly all of M222's history has been bottlenecks, briefly broken by only a few explosive growth spurts. Is this what you believe?

Dubhthach
09-07-2015, 03:48 PM
I don't deny that rapid growth can occur. But assuming only randomly distributed single SNPs requires you to not only believe that a growth spurt occurred. It requires you to believe that nearly all the growth occurred within brief spurts of just a few generations - so that there was no time for any SNPs to break a multibranch node - and then there was NO significant growth for a long time, allowing a block of equivalent to SNPs to form, and then another brief spurt, and another drought, etc. Assuming evenly distributed single SNPs, the only compatible interpretation is that nearly all the growth occurred through just a few brief spurts, with nothing in between.

It defies belief. And there is no reason to embrace such a difficult position when there is already ample evidence that the SNPs are NOT evenly distributed.

Tbh I don't think I'm arguing for SNP's to be evenly distributed, as mentioned before there hasn't been any detailed studies into level of new snp acquistion in an extended lineage. Leaving aside the mass block of 42 snp's at M222 equivalent (going off BigY/NGS), I don't see any other comparable blocks within the downstream clades. This is reflective of fact that if we are to believe the traditional genealogical account (as outlined for example in Dubhaltach Mac Fhirbhisigh magnus opus of the 1650's) that the lineages underwent prolonged and contuined expansion for well over 1000 years, basically running from 5th century until the destruction of Gaelic Ireland during the Tudor conquest of Ireland in period after 1535.

What does the traditional account tell us. One that the Uí Néill descend from multiple sons of Níall Noígiallach whose floruit is put in late 4th/early 5th century (Historians would argue for mid 5th century), that likewise the three Connachta descend from his half-brothers. From point of view of the traditional lineages, a huge amount of surnames/dynastically groups collapse down to one man in this case their father Eochaid Mugmedón.

From this man within the genealogical tradition there spawns a total of 10-15 dynastical groups, all of which segregate from each other over period of less than 200 years. From a surname point of view you are looking at perhaps at least 100-200+ distinct surnames falling into these groups.

Now is this plausible situation? Probably not, genealogies were subject to manipulation particular for political means. What's interesting though is how robust some of them are holding up.

For example under A260 we have Follan and Reilly, if we believe the traditional account (as outlined by Mac Fhirbsigh), the two lineages would have separated from each other in mid 6th century. Likewise McHugh (if we are talking about Uí Briúin Seola McHugh) likewise spilt at same time in genealogical sources. Between the three we are seeing only a share block of two SNP's (A260 and 20816920-G-A). The GD's are of the order of up to 16 (McHugh <-> Fallon) at 67 markers, which is massive in terms of both M222, but also within a specific subclade within M222.

Mac von Frankfurt
09-07-2015, 03:54 PM
I need a little help understanding this block or cluster idea. How do we tell the difference between five mutations occurring over one generation (between a father and son A but not son B ) and five mutations occurring down the line of son A at regularly spaced intervals over 500 years, but there is only one surviving descendant of son A at the end of that 500 years. In other words, if we observe a block how do we know if the cause is a) a high mutation rate, b) no survival of intermediate branches, or c) a lack of data.

Dubhthach
09-07-2015, 03:57 PM
Now what would be nice would be to do BigY/FGC on both the "O'Conor Don" and "the McDermott" these two men have unbroken lineage right back. With the two lines seperating from each other by been descended from different sons of Tadg mac Cathail who died in 925.

miiser
09-07-2015, 04:03 PM
Tbh I don't think I'm arguing for SNP's to be evenly distributed, as mentioned before there hasn't been any detailed studies into level of new snp acquistion in an extended lineage. Leaving aside the mass block of 42 snp's at M222 equivalent (going off BigY/NGS), I don't see any other comparable blocks within the downstream clades. This is reflective of fact that if we are to believe the traditional genealogical account (as outlined for example in Dubhaltach Mac Fhirbhisigh magnus opus of the 1650's) that the lineages underwent prolonged and contuined expansion for well over 1000 years, basically running from 5th century until the destruction of Gaelic Ireland during the Tudor conquest of Ireland in period after 1535.

The belief that SNPs are evenly distributed is implicit in your assumption that the number of branches per node and number of equivalent SNPs per node correlates strongly with the growth rate of the lineage. Once one accepts that SNPs sometimes occur in runs, with longer intervening intervals devoid of any SNPs, it is no longer reasonable to make this assumption. The absence of large equivalent blocks further downstream may have nothing to do with the lineage growth rate, but may just be a random outcome of the absence of large runs.

Dubhthach
09-07-2015, 04:22 PM
The belief that SNPs occur are evenly distributed is implicit in your assumption that the number of branches per node corresponds strongly with the growth rate of the lineage. Once one accepts that SNPs sometimes occur in runs, with longer intervening intervals devoid of any SNPs, it is no longer reasonable to make this assumption.

But we have plenty of documentary evidence that the lineages did undergo rapid expansion, even right up until 15th century, you are seeing multiple examples of ruling dynasts producing upwards of 40 grandsons, that is matter of fact, you can check the English state papers for Ireland if you don't believe it. if that's not rapid expansion I don't know what is.

I don't think I've stated anywhere that one SNP could only arise per generation. So in case of comparing a A259+ man and a DF85+ man, I don't have any issue with thinking that the two SNP's the share with each other (DF105 and DF109) might have arisen in just one man (eg. both SNP's mutated when he was conceived).

Likewise it wouldn't surprise me if 20816920-G-A arose at same time as A260 (in which case to be A260+ means you are automatically 20816920-G-A) under A259, however we know that with at least one A259+ man that he only shares one snp (A259) with the A260+/20816920-G-A men. (the other A259+/A260- men haven't done BigY so we don't know status of 20816920-G-A).

Until we can say we can test all lineages, we'll never know how the SNP's fully spilt out, of course there is also the issue with regards to coverage in FGC vs. BigY. I'm not sure if on some of these clades under DF105 if we have seen at least 2 men doing FGC testing. So it's quite possible that some of the blocks that are currently only 1 or 2 snp's are larger.

Dubhthach
09-07-2015, 04:37 PM
I need a little help understanding this block or cluster idea. How do we tell the difference between five mutations occurring over one generation (between a father and son A but not son B ) and five mutations occurring down the line of son A at regularly spaced intervals over 500 years, but there is only one surviving descendant of son A at the end of that 500 years. In other words, if we observe a block how do we know if the cause is a) a high mutation rate, b) no survival of intermediate branches, or c) a lack of data.

It's fair enough point, to be honest there might be dozens's of branches/lineages from such a scenario and you might only have 2 or 3 tested. When it comes to understanding the source of M222 I think of lack of data might be major issue. There was some assumptions that M222 in Northern Britain was older based on STR variance, there is also at least one Scottish man who is M222+/S7073- (all other M222+ men tested are S7073+), however S7073 seems to be unreliable.

What we can tell looking at parallel clades under Z2961 (which is showing up as a block of 2 snp's) that some of these clusters are in Britain, the other thing is that majority I would say of Irish origin M222 that have undergone BigY/FGC or M222 bundle appears to fall under DF105 (from brief observation, looking at surnames of Gaelic Irish origin), which is 3-4 "steps" down from M222 "super-block". The question than is when we look at distribution of M222 within both Ireland and Britain what percentage of all M222 is DF105+ in Ireland and how does that compare to say Scotland/England etc.

miiser
09-07-2015, 05:14 PM
But we have plenty of documentary evidence that the lineages did undergo rapid expansion, even right up until 15th century, you are seeing multiple examples of ruling dynasts producing upwards of 40 grandsons, that is matter of fact, you can check the English state papers for Ireland if you don't believe it. if that's not rapid expansion I don't know what is.

I don't think I've stated anywhere that one SNP could only arise per generation. So in case of comparing a A259+ man and a DF85+ man, I don't have any issue with thinking that the two SNP's the share with each other (DF105 and DF109) might have arisen in just one man (eg. both SNP's mutated when he was conceived).

Likewise it wouldn't surprise me if 20816920-G-A arose at same time as A260 (in which case to be A260+ means you are automatically 20816920-G-A) under A259, however we know that with at least one A259+ man that he only shares one snp (A259) with the A260+/20816920-G-A men. (the other A259+/A260- men haven't done BigY so we don't know status of 20816920-G-A).

Until we can say we can test all lineages, we'll never know how the SNP's fully spilt out, of course there is also the issue with regards to coverage in FGC vs. BigY. I'm not sure if on some of these clades under DF105 if we have seen at least 2 men doing FGC testing. So it's quite possible that some of the blocks that are currently only 1 or 2 snp's are larger.

Okay, you're apparently not understanding my argument, because you're not responding to the argument that I'm making. But I'm not sure how else to put it to make it any clearer, so I'm afraid the discussion may have reached a failed end. But I'll give it the old college try one more time before I throw in the towel.

You appear to be arguing so: because there are not such large numbers of equivalent SNPs per node further down in the tree, therefore the lineage must have been growing more rapidly during this time.

I am arguing that this is an invalid line of reasoning, because of the fact that SNPs are unevenly distributed. The reality may be that the growth rate was the same at the top and bottom of the tree, but there just happened to be a quick run of many SNPs at the top of the tree.

The contradiction between what I'm saying and what you're saying is not that you're saying the growth rate was fast and I'm saying it was slow. The contradiction is that you are saying the difference in the tree structure between the top and bottom is an indication of a change in growth rate. And I'm saying that the change in structure tells us nothing about the growth rate, but is probably just due to random chance that the top of the tree happened to experience a rapid, very big SNP run. In other words, the number of branches per node and the number of equivalents per node cannot indicate what the growth rate was, because the number of branches per node and the number of equivalents per node have been distorted by the uneven SNP distribution.

I think there may be no way to convince you unless you actually go through the exercise of making a population growth model for yourself to see what I'm talking about. It doesn't even have to be a fancy computer model. You can do it on a piece of paper. Just make a tree, and draw some reasonable number of sons (with surviving lineages) per man, and then randomly mark every 1 out of 4 with a make believe SNP. Do the same thing the next generation, and again, and again, etc. If you look at the bottom of the tree, you will find that most of the nodes have just 2 branches and at most a few equivalent SNPs.

But the real life M222 tree, on average, has quite a bit more than 2 branches per node and a few SNPs per node. Using a reasonable average mutation rate in line with what we've observed, there's no way to match the structure we actually see with a model . . . UNLESS you squish a bunch of the SNPs together into a packet that occurred instantaneously. So you end up with a big packet here, a couple small packets here and there, and still a bunch of single SNPs, but spaced out every ~500 years instead of every ~140 or whatever.

And the result of this very uneven distribution of SNPs is that the number of branches per node and the number of equivalents per node no longer tell you anything about how much growth the lineage was experiencing at any given time.

If the SNPs occur at an even pace, we can use the node size and equivalent block size to make reasonable deductions about the growth rate of the lineage. If the SNPs are scrunched together here and there, with multiple SNPs sometimes occurring instantaneously (approximately), then this obscures any information the tree might have contained about growth rate, had the SNPs been evenly distributed. So we are no longer able to make deductions about bottle necks, growth spurts, etc. based on the tree structure.

The problem I have with your argument is that you are trying to deduce growth rate based on node size and equivalent block size. But, due to the existence of runs, this is an invalid line of reasoning.

miiser
09-07-2015, 05:18 PM
I need a little help understanding this block or cluster idea. How do we tell the difference between five mutations occurring over one generation (between a father and son A but not son B ) and five mutations occurring down the line of son A at regularly spaced intervals over 500 years, but there is only one surviving descendant of son A at the end of that 500 years. In other words, if we observe a block how do we know if the cause is a) a high mutation rate, b) no survival of intermediate branches, or c) a lack of data.

YOU GET IT!

This is exactly the problem. We don't know. So this makes it impossible to deduce, when we see a large block of equivalent SNPs or a node with multiple branches, that there's been a population bottleneck or a growth spurt.

We have NO IDEA when each of those mutations actually occurred, so we can't assume we know when the branch occurred.

Over long time periods, by averaging out SNP counts through multiple parallel branches, the "jerkiness" of the data smooths out so that you can make reasonable estimations of how long it took for a total number of SNPs to occur. But if we see 10 mutations in an equivalent block at a node, we have no information to tell us whether they occurred in 1 generation or 50 generations.

In general, if the SNPs are spread out pretty evenly through the tree as single SNPs, then you will end up with a tree that has mostly 2 branches per node and usually 1 or a few SNPs per node. If the SNPs tend to be bunched together into groups you will end up with a tree that has more branches per node and more SNPs per block. The tree that we actually have matches the second scenario.

The effect of this is that we can't tell whether a branch was growing quickly or slowly based on how many branches and equivalents it has at a given node, because we don't know whether the SNPs were bunched together into a group or spread out evenly.

Rory Cain
09-07-2015, 08:30 PM
[QUOTE=Dubhthach;107300]The issue we have at moment is that ScotlandsDNA/IrelandsDNA hasn't release any data with regards to division of various M222 subclades within their testing database...
...They did release a generic map, but it seems to be just all of M222...

http://compsoc.nuigalway.ie/~dubhthach/DNA/m222_spread.png

[QUOTE]

Thanks, Paul. Notwithstanding your comments re lack of sub-clades, which would add another layer to our understanding, there is another valuable use for this map. It could be a useful exercise to compare it with say DF21 and DF1/L513.

Predictions are frought with danger but I anticipate that we would find both similarities and differences. Similarities are likely to include concentrations in Southern Scotland and South Ulster where all M222, DF21 and L513 cluster in some sort of unison. But I expect there are also differences In distribution. I am less confident that these have received equal attention.

Dubhthach
09-08-2015, 09:46 AM
Well you have to remember maps like above were created only for the select haplogroups that ScotlandsDNA decided to give "marketing speel" to. I don't know if they created a "profile" for DF21 or L513, if they didn't than it's probable they role them up into their generic L21 map

https://dnaexplained.files.wordpress.com/2014/07/s145.jpg

When you combine the percentages for L21(-m222) and M222 you see following


Ulster -- 53%
Connacht -- 67%
Leinster -- 60%
Munster -- 61%
Scotland South West -- 43%


These percentages at least come closer to the figures as outlined in Busby (2011) for total L21 percentage in any given region of Ireland.

It wouldn't surprise me if they have maps for L159.2 and L226 as I believe they have "speels" for those (Dalcassian and Hibernian)

Dubhthach
09-08-2015, 10:44 AM
With regards to subclades of M222 within Ireland Rory, I had a look in Ireland project at men who have results for the M222 bundle test. The bundle tests for the following:
http://compsoc.nuigalway.ie/~dubhthach/DNA/ftdna-m222.png

Obviously it's early days so far, but we have results for 128 men these break out as following
n=128

M222* = 0.78%
FGC4077+ = 13.28%
S568+ = 2.34%
S658+ = 83.59%


The S658+ is made up of n=107

DF104+/DF105-: 0.93% -- 0.78% of total sample (n=128)
DF105+: 99.07% (82.81% of total sample)


Good possibility that the DF104+/DF105- needs to be checked to make sure there's no issue with this result

DF105+ (n=106) makes up 82.81% of total sample. It appears to break into following subclades:

S588+: 19.81% (n=21, 16.41% of total sample)
DF85+: 24.53% (n=26, 20.31% of total sample)
A260+: 20.75% (n=22, 17.19% of total sample)
BY198+: 4.72% (n=5, 3.91% of total sample)
A223+: 2.83% (n=3, 2.34% of total sample)
FGC4133+: 0.94% (n=1, 0.78% of total sample)
DF105*: 26.42% (n=28, 21.88% of total sample)


Part of issue I have with M222 bundle is it doesn't include A259, when it was designed A259 was assumed to be equivalent (in block) with A260, so they only picked one of two SNP's. We've since seen via SNP testing and BigY that A259 is upstream of A260. So there possibility that some of the DF105+ are A259+/A260-. We know that one of men has also done BigY and is A1206+, this is fairly new SNP under DF105 found in three BigY testees. Without it in bundle it's hard to say how much if any dent it would put into the DF105 group.

What's probably needed is a number of targeted BigY test's.

S588, DF85 and A259 (in shape of A260+) make up 65.09% of the DF105 sample (53.19% of total sample). It would be good to know what their distribution in M222 elsewhere was like. I believe both S588 and DF85 are on the Chromo2 test (neither A259 or A260 are), so if data was made public you could at least plot the differences between various geographic regions they've drawn on Ireland/Britain map.

Given what appears to be dominance of DF105 within the Irish tested M222 so far I be curious if dynamic is different in Scotland/Northern England.

Rory Cain
09-08-2015, 08:30 PM
...Given what appears to be dominance of DF105 within the Irish tested M222 so far I be curious if dynamic is different in Scotland/Northern England.

There are several septs named O'Cathain, unrelated except for most being R-L21. Most folks are familiar only with one - the O'Cathain of Ciannachta Glinne-Geimhin or modern day North Co Derry. They received a little attention when it was discovered that most Co Tyrone O'Neills are not the so-called "Niall of the Nine Hostages" DNA signature", but other Septs descended from Niall are, including O'Cathain & MacLoughlin.

Further testing makes the O'Cathains of Derry R-S588, but few of them have tested for it. Perhaps they are satisfied with the "15 seconds of fame" they enjoyed when the scientific community took a brief interest in them. To conclude, I share your thoughts. When we are so close, would be nice to have the evidence wrapped up.

Dubhthach
09-09-2015, 10:27 AM
There are several septs named O'Cathain, unrelated except for most being R-L21. Most folks are familiar only with one - the O'Cathain of Ciannachta Glinne-Geimhin or modern day North Co Derry. They received a little attention when it was discovered that most Co Tyrone O'Neills are not the so-called "Niall of the Nine Hostages" DNA signature", but other Septs descended from Niall are, including O'Cathain & MacLoughlin.

Further testing makes the O'Cathains of Derry R-S588, but few of them have tested for it. Perhaps they are satisfied with the "15 seconds of fame" they enjoyed when the scientific community took a brief interest in them. To conclude, I share your thoughts. When we are so close, would be nice to have the evidence wrapped up.

Well there are at least two S588+ O'Neill's one of them actually from Dungannon, other one just done a BigY so it be interesting to see how he compares against other S588+ BigY's (I think there's about 20 of them tested to BigY so far). The S588+ men who've done the bundle break down as following:

n=21

F499+ - 1 (4.76%)
S7049+: 3 (14.29% -- rounded)
S7814+ - 2 (9.52%)
S603+ - 3 (caveat not included in bundle, SNP tested) (14.29% rounded)


The leaving out of S603 from bundle looks like a large mistake.

If we look at DF85 we get a breakdown as follows
n=26

DF85 (26)

S673 (22)

CTS3771 (0)
S668 (19)

CTS12173 (0)
DF97/FGC8739 (15)

CTS8002 (0)
M226 (0)
CTS6 (0)



DF97/FGC8739 (equivalents) make up 57.69% of sample, there's total of 4 DF85* (15.38%), 3 S673* (11.53%) and 4 S668* (15.38%).

A259 wasn't really that well served in bundle, not really surprising as it was quite new. We have to use it's subclade A260 as a proxy. Of the 22 A260+ 11 fall into FGC5939+ (50%), they didn't include A883 which is unfortunate given how it seems to show up in certain Bréifne surnames.

George Chandler
09-10-2015, 05:16 AM
I understand that SNP test methods vary, and this is why I use the term "observed" mutations when I discuss such analysis. My analysis does not assume that every SNP is detected. It assumes that a Big Y test method is used, which is true for nearly all the available test samples. There will be some variation in the number of base pairs sequenced and the coverage from one test to another. This difference is not able to account for the varying number of SNPs seen from one branch to another. The difference of SNPs from one branch to another is a difference of SNPs observed per BP sequenced. The method used is standardized, and, on average, the fraction of detected SNPs should be similar from one sample to another throughout the tree in different branches. No matter how you do the analysis, if you do an apples to apples comparison, there is excessive variation in the count beyond what is expected for single, random mutations, beyond what can be accounted for by differences in test method.

The test method and missed SNPs is wholly unable to account for large blocks of equivalent SNPs. Test method is not a factor here.

You ask whether I am "able to capture multiple Y SNP mutations in a single person's sample which are reliable". There are numerous people who have multiple novel, private SNPs with good confidence from Big Y, if that is what you are asking for. If you are asking for a measure of the number of SNPs between father and son, the answer is that we don't have the data because very few people find a reason to purchase a NGS test for both father and son.

I agree that there have been bottlenecks. I also believe that bottlenecks and spurts are the exception, not the norm. Believing that SNPs occur as single, randomly distributed events does not just require you to believe that there have been bottlenecks. It requires you to believe that nearly all of M222's history has been bottlenecks, briefly broken by only a few explosive growth spurts. Is this what you believe?

I think I understand what where you're coming from but if I'm missing your point let me know - don't you think the more likely explanation for the numbers (or "blocks" of equivalent SNP's) is proportional to the number of people testing that specific lineage? I agree there is the possibility of natural runs and gaps but the more likely cause (IMO) is that first..those testing are not targeting the specific lineage enough to try and reduce the number of observable existing equivalent SNP's..secondly SNP's removed due to reliability issues or thirdly lineages which daughter out or go extinct? You will get a series of SNP's that you will never be able to position just because lines have gone extinct.


I know you say test method is not a factor here if you're using the same testing parameters in you're research (though can't account for the methods you're using). Some people test Big Y or Y Elite and post everything (all observed SNP's). Some post only reliable SNP's..some post only high quality SNP's..some double verify the reliable SNP's through a repeat test and some do not. Some test known genealogical lines and some don't.

I haven't seen enough of the M222 results to determine what lines may have been bottlenecked or not. I think there should be something historical such as a natural disaster, plague etc which connects the culture and the estimated SNP age.

George

miiser
09-10-2015, 06:17 AM
I think I understand what where you're coming from but if I'm missing your point let me know - don't you think the more likely explanation for the numbers (or "blocks" of equivalent SNP's) is proportional to the number of people testing that specific lineage? I agree there is the possibility of natural runs and gaps but the more likely cause (IMO) is that first..those testing are not targeting the specific lineage enough to try and reduce the number of observable existing equivalent SNP's

It is true that increased sampling will break up some equivalent blocks into separate branches. But, keep in mind, I argued that the evidence for runs lies in the number of branches per node and the excessive variation of the SNP count in sibling branches. New samples are not able to reduce the number of branches per node. They can only create new nodes or add more branches to an already discovered node. So the large number of branches per node, which is already greater than what is consistent with random, single SNPs, can only possibly grow larger with additional samples.

Likewise, the excessive branch-to-branch variation of SNP count will not be affected by additional samples.

So, even if increased sampling reduces the number of equivalents, we still have to explain the high number of branches per node and the excessive branch-to-branch variation of SNP count. And this can only be explained by SNP runs.


..secondly SNP's removed due to reliability issues or thirdly lineages which daughter out or go extinct? You will get a series of SNP's that you will never be able to position just because lines have gone extinct.

If you make a population model and study how the extinct branches affect the structure of the modern, surviving branches, you will find that you can basically ignore the extinct branches. If a father might have had 8 sons, with 2 lineages surviving to modernity, you can instead just give him 2 children and ignore the 6 extinct branches. They don't affect the final structure. If there is one SNP in every 4 children, it will still be one SNP in every 4 children whether or not a particular lineage survives. So it doesn't affect the final structure of the tree. It is only the percentage of offspring containing an SNP which will determine the structure of the final tree.

I will make the same suggestion to you that I've made to others. Go ahead and play around with making some simple models on paper. Some of these behaviors are not obvious, and they may not make sense to you until you actually tinker with a model yourself.


I know you say test method is not a factor here if you're using the same testing parameters in you're research (though can't account for the methods you're using). Some people test Big Y or Y Elite and post everything (all observed SNP's). Some post only reliable SNP's..some post only high quality SNP's..some double verify the reliable SNP's through a repeat test and some do not. Some test known genealogical lines and some don't.

I've compared the raw data of sibling branches in a variety of ways, and they all consistently indicate an excessive degree of SNP count variation, beyond what is expected from random, single SNPs. Basically, in order to do an apples to apples comparison, you look at the raw data files and compare two parallel branches from the same node, over the same time range, from a terminal non-private SNP up to the present time. (There are plenty of raw data files uploaded to Yahoo groups and such.) You only compare samples that were tested by the same method (Big Y, because that's what we have the most data for), and you look within the same region on both samples, and you only do this with samples that have comparable coverage in that region. And then you compare the number of private SNPs called at a specific confidence level. There are a variety of ways you can do an apples to apples comparison. But no matter how you do it, you will find that the variation far exceeds what is consistent with single, random SNPs.


I haven't seen enough of the M222 results to determine what lines may have been bottlenecked or not. I think there should be something historical such as a natural disaster, plague etc which connects the culture and the estimated SNP age.

If there is a known event which might cause a bottleneck, and this effect is observed across multiple branches in the region, then it is reasonable to suppose that there may be a true bottleneck. If the equivalent SNP block is seen in one haplogroup only, but all the other haplogroups residing in the same area don't similarly exhibit equivalent blocks at around the same time, then it is unlikely to be a population effect that selectively targeted only a single haplogroup and ignored all the others in the neighborhood. Natural disasters, plagues, and such are not able to do DNA sequencing and cannot tell one haplogroup from another.

And, again, I would point out that the structure of the tree, if it's assumed to be made up of single, random SNPs, requires nearly an entire history of mostly bottlenecks. For many of the haplogroups, you have a major fraction of the total branches occurring within 2 or 3 nodes over thousands of years. This implies that there were continuous bottlenecks, broken by only a few brief, rapid growth spurts of 1 to a few generations. That is, unless you allow that some of the SNPs may have occurred as runs and SNPs do not create a new branch every 4 or 5 generations.

I really do encourage you to fiddle with some models to see how this plays out. Single, random SNPs just aren't capable of producing the tree that we observe.

George Chandler
09-11-2015, 05:24 AM
because lines have gone extinct.
If you make a population model and study how the extinct branches affect the structure of the modern, surviving branches, you will find that you can basically ignore the extinct branches. If a father might have had 8 sons, with 2 lineages surviving to modernity, you can instead just give him 2 children and ignore the 6 extinct branches. They don't affect the final structure. If there is one SNP in every 4 children, it will still be one SNP in every 4 children whether or not a particular lineage survives. So it doesn't affect the final structure of the tree. It is only the percentage of offspring containing an SNP which will determine the structure of the final tree.


What I'm thinking here is how those extinct lines factor into trying to position those blocks of equivalent SNP's. You're right it won't affect the final structure in one sense, but it just makes it difficult to place them. If I have a block of four SNP's from one test (say 1,2,3 & 4) and the next set of results has SNP 4 but not 1,2 &3. So obviously I can place 4 as the oldest but now the historical branches which have SNP's 4 and say 3 as then next oldest have all gone extinct so I'm left with an unknown as to which is the next in line.


If there is a known event which might cause a bottleneck, and this effect is observed across multiple branches in the region, then it is reasonable to suppose that there may be a true bottleneck. If the equivalent SNP block is seen in one haplogroup only, but all the other haplogroups residing in the same area don't similarly exhibit equivalent blocks at around the same time, then it is unlikely to be a population effect that selectively targeted only a single haplogroup and ignored all the others in the neighborhood. Natural disasters, plagues, and such are not able to do DNA sequencing and cannot tell one haplogroup from another.


You can have genetic bottlenecks though which don't have to be largely geographic. Another form can be cultural which can take the form of genocide directed at a certain ethnic or tribal group. It may have little or no affect on the genetics of the surrounding populations but will genetically bottleneck a certain tribe or ethnic group within. Something else to consider is how large the sample size is when the bottleneck happens. So let's say it's L21 found within only 1,000 people in the ancient world and during that time period and war breaks out and you lose 90% of the population you're now left with 100 people who carry L21. Compare that with 100,000 people who carry L21 and then losing 90% of the population due to war you're left with 10,000 people carrying L21. The 1,000 people may have struggled through hardship over a 1,000 year period just to get their numbers to 1,000 people. So in theory you could have say 5 lines of 5 new SNP's below L21 at the time for those 1,000 people then after the 90% are lost you are left with a line of five single SNP's. Now if you look at the scenario of 100,000 people the number of unique SNP's an lineages post bottleneck will likely be totally different.

I do think you're correct that you can find those natural runs and gaps but I just think it's probably the less likely scenario for what you're observing (IMO). I understand what you're saying but looking at historical events such as wars, genocide, famine etc it's a wonder to me that anyone survived. In ancient times they never had old age pension so if you needed sons to work the fields and to look after you in your ancient golden years then you had a lot of them (or tried to). So if a tribe or region was decimated by a plague and you lose 90% of the population then what do you do?..have as many kids as fast as possible to replace those you've lost. I'll try and take a closer look to see how it plays out when I have some spare time.

George

Dubhthach
09-11-2015, 09:01 AM
I haven't seen enough of the M222 results to determine what lines may have been bottlenecked or not. I think there should be something historical such as a natural disaster, plague etc which connects the culture and the estimated SNP age.

George

If we are to accept the supposition that M222 is somehow linked to the lineages of the Dál Cuinn in Ireland (eg. not that M222 arose among them, but that M222 was one of major lineages in grouping), than going off historical record you at least have a point where such lineages enter exponential growth. The historical research into various Dál Cuinn lineages (Uí Néill and the three Connachta -- Uí Briúin, Uí Fiachrach, Uí nAilleo) point to widespread growth of lineages and a basic "overthrow of the old order" in period that Christianity is introduced into Ireland (5th century onwards).

What's seems to be evident is that in period of 200 years or so we see rapid expansion, to point where come the 7th century there is a general rewriting of Irish history by what modern historians call the "Synchronistic historians". Which basically back-write the pervailing political situation of the 7th-8th century back into deep antiquity. The same of course can be said for expansion of Eoghanachta in Munster (which some have linked to CTS4466).

At the same time the Irish language undergoes rapid change, which is fairly unprecedented, the shift from Archaic Irish of Ogham to Old Irish (600-900AD) is fairly massive, on order of shift from Classic Latin to say Old French. Now there's some argument that "Archaic Irish" usage in the 4th and 5th century on Ogham was perhaps retention of an "archaic standard" for ritualistic purposes. What's evident though is that Irish society is massively remodelled from what appears to be "Tribal society" to one obsessed about lineages.

So much so that Mac Fhirbsigh for example would write in the 17th century disparaging about mere Churls who don't even know who their great-grandfather was!

miiser
09-11-2015, 11:20 PM
What I'm thinking here is how those extinct lines factor into trying to position those blocks of equivalent SNP's. You're right it won't affect the final structure in one sense, but it just makes it difficult to place them. If I have a block of four SNP's from one test (say 1,2,3 & 4) and the next set of results has SNP 4 but not 1,2 &3. So obviously I can place 4 as the oldest but now the historical branches which have SNP's 4 and say 3 as then next oldest have all gone extinct so I'm left with an unknown as to which is the next in line.



You can have genetic bottlenecks though which don't have to be largely geographic. Another form can be cultural which can take the form of genocide directed at a certain ethnic or tribal group. It may have little or no affect on the genetics of the surrounding populations but will genetically bottleneck a certain tribe or ethnic group within. Something else to consider is how large the sample size is when the bottleneck happens. So let's say it's L21 found within only 1,000 people in the ancient world and during that time period and war breaks out and you lose 90% of the population you're now left with 100 people who carry L21. Compare that with 100,000 people who carry L21 and then losing 90% of the population due to war you're left with 10,000 people carrying L21. The 1,000 people may have struggled through hardship over a 1,000 year period just to get their numbers to 1,000 people. So in theory you could have say 5 lines of 5 new SNP's below L21 at the time for those 1,000 people then after the 90% are lost you are left with a line of five single SNP's. Now if you look at the scenario of 100,000 people the number of unique SNP's an lineages post bottleneck will likely be totally different.

I do think you're correct that you can find those natural runs and gaps but I just think it's probably the less likely scenario for what you're observing (IMO). I understand what you're saying but looking at historical events such as wars, genocide, famine etc it's a wonder to me that anyone survived. In ancient times they never had old age pension so if you needed sons to work the fields and to look after you in your ancient golden years then you had a lot of them (or tried to). So if a tribe or region was decimated by a plague and you lose 90% of the population then what do you do?..have as many kids as fast as possible to replace those you've lost. I'll try and take a closer look to see how it plays out when I have some spare time.

George

I'm getting worn out by this discussion, so I'm going to let it end here and get back to my real world life. But there's just one more observation, which I hadn't previously mentioned, that I need to point out. It appears to be the case that some people believe that population bottlenecks will result in a long lineage without any branching, with many phylogenetically equivalent SNPs. But this is NOT in fact what typically happens in a population bottleneck.

For example, suppose there is a small haplogroup of only 20 males. Suppose this haplogroup, for whatever reason, experiences no population growth for 1000 years. And so, at the end of 1000 years, this haplogroup is still limited to only 20 males. One might suppose that these 20 males are all going to be from the same branch after such a long bottleneck, but this couldn't be further from the truth. More likely, you will end up with 20 males from 20 different branches, or nearly 20 different branches. Whatever growth or culling occurs, it will not leave just one lineage surviving in each generation. The growth and culling will be somewhat randomly distributed between each of the 20 lineages in each generation. You will have one lineage surviving through one son, and a different lineage surviving through another son. The surviving branches in each generation are not going to be constrained to just a single lineage. Even though the total population doesn't increase, every so often you will have another branch appearing in a sibling or cousin. Even a very small population will be periodically bifurcated by a new SNP. And so that population of 20 will over time gradually be fragmented into pieces, from one branch at the beginning down to 20, or close to 20 branches, at the end. Even when the population doesn't increase over time, the number of branches still does.

Unless a haplogroup experiences a bottleneck which cuts it down to literally just a SINGLE individual, there will still be multiple branches containing different SNPs, even in a very small population. And hypothetical bottlenecks which constrain a haplogroup population to just a single individual without extinguishing it entirely are so remotely improbable that they are not even worth considering as a reasonable explanation for an SNP block.

Even when population bottlenecks do occur, they don't have the effect of creating one single, continuous, homogeneous lineage with numerous equivalent SNPs. Branching still occurs even in a population which is experiencing no growth or shrinkage. So even if many extended duration bottlenecks have occurred in all the major haplogroups, as some people suppose they have, they would not have created long stringy lineages without branching, with numerous equivalent SNPs. Only large SNP runs occurring in a short time are capable of producing such features.

jdean
09-12-2015, 12:36 AM
I'm getting worn out by this discussion, so I'm going to let it end here and get back to my real world life. But there's just one more observation, which I hadn't previously mentioned, that I need to point out. It appears to be the case that some people believe that population bottlenecks will result in a long lineage without any branching, with many phylogenetically equivalent SNPs. But this is NOT in fact what typically happens in a population bottleneck.

For example, suppose there is a small haplogroup of only 20 males. Suppose this haplogroup, for whatever reason, experiences no population growth for 1000 years. And so, at the end of 1000 years, this haplogroup is still limited to only 20 males. One might suppose that these 20 males are all going to be from the same branch after such a long bottleneck, but this couldn't be further from the truth. More likely, you will end up with 20 males from 20 different branches, or nearly 20 different branches. Whatever growth or culling occurs, it will not leave just one lineage surviving in each generation. The growth and culling will be somewhat randomly distributed between each of the 20 lineages in each generation. You will have one lineage surviving through one son, and a different lineage surviving through another son. The surviving branches in each generation are not going to be constrained to just a single lineage. Even though the total population doesn't increase, every so often you will have another branch appearing in a sibling or cousin. Even a very small population will be periodically bifurcated by a new SNP. And so that population of 20 will over time gradually be fragmented into pieces, from one branch at the beginning down to 20, or close to 20 branches, at the end. Even when the population doesn't increase over time, the number of branches still does.

Unless a haplogroup experiences a bottleneck which cuts it down to literally just a SINGLE individual, there will still be multiple branches containing different SNPs, even in a very small population. And hypothetical bottlenecks which constrain a haplogroup population to just a single individual without extinguishing it entirely are so remotely improbable that they are not even worth considering as a reasonable explanation for an SNP block.

Even when population bottlenecks do occur, they don't have the effect of creating one single, continuous, homogeneous lineage with numerous equivalent SNPs. Branching still occurs even in a population which is experiencing no growth or shrinkage. So even if many extended duration bottlenecks have occurred in all the major haplogroups, as some people suppose they have, they would not have created long stringy lineages without branching, with numerous equivalent SNPs. Only large SNP runs occurring in a short time are capable of producing such features.

Perhaps you could provide a reasonable modal demonstrating this ?

Dubhthach
09-12-2015, 01:33 AM
Of course perhaps a good quote about origins of M222, would be to go with late John V. Kelleher remark about the Uí Néill


The Uí Néill emerge into history like a school of cuttlefish from a large ink-cloud of their own manufacture; and clouds and ink continued to be manufactured by them or for them throughout their long career. Only one thing seem consistent, their claim of sole right to the kingship of Tara.

As an aside I collected results from the M222 bundle for men who had tested in M222 project (scraped their SNP results page), obviously a good chunk of these are in Ireland project as well, though numbers are higher


$ wc -l M222-project.txt
217 M222-project.txt



M222* = 2 (0.92%)
FGC4077+ = 24 (11.06%)
S568+ = 6 (2.77%)
S658+ = 185 (85.25%)


The S658+ breaks down as following:

DF105- = 4 (2.16%)
DF105+ = 181 (97.84%) -- 83.41% of total bundle


The DF105+ in itself breaks down into:

S588+ = 34 (18.78%) -- 15.67% of total bundle
DF85+ = 50 (27.62%) -- 23.04% of total bundle
A260+ = 30 (16.57%) -- 13.82% of total bundle
BY198+ = 9 (4.97%) -- 4.14% of total bundle
A223+ = 12 (6.63%) -- 5.53% of total bundle
FGC4133+ = 5 (2.76%) -- 2.3% of total bundle
DF105* = 41 (22.61%) -- 18.89% of total bundle.



With just less than double the sample from early overview the picture is amazingly consistent. The three major branches know branches of DF105, which appear to map onto a number of historic kindreds (on surname analysis) make up over 50% of sample. Given the importance of these 3 specific kindreds in Irish history, it's not surprising that if a SNP was linked to one that it could grow to large size within the overall M222 population. From a traditional genealogical picture these lineages spilt at beginning of the historic expansion of Dál Cuinn power specifically under what would later be known as the Uí Néill. A process which would lead to 50-60% of Ireland falling under their political hegemony within a period of about 200 years.

Megalophias
09-12-2015, 04:43 PM
For example, suppose there is a small haplogroup of only 20 males. Suppose this haplogroup, for whatever reason, experiences no population growth for 1000 years. And so, at the end of 1000 years, this haplogroup is still limited to only 20 males. One might suppose that these 20 males are all going to be from the same branch after such a long bottleneck, but this couldn't be further from the truth. More likely, you will end up with 20 males from 20 different branches, or nearly 20 different branches. Whatever growth or culling occurs, it will not leave just one lineage surviving in each generation. The growth and culling will be somewhat randomly distributed between each of the 20 lineages in each generation. You will have one lineage surviving through one son, and a different lineage surviving through another son. The surviving branches in each generation are not going to be constrained to just a single lineage. Even though the total population doesn't increase, every so often you will have another branch appearing in a sibling or cousin. Even a very small population will be periodically bifurcated by a new SNP. And so that population of 20 will over time gradually be fragmented into pieces, from one branch at the beginning down to 20, or close to 20 branches, at the end. Even when the population doesn't increase over time, the number of branches still does....

New branches form and old branches die out. Say you have a small population founded by 20 men, representing 20 lineages. In each generation some of those men will have no children, some will have only daughters, some will have just one son, some will have many sons. In the next generation, if the population does not grow, there will still be the same number of men, but now some of the original lineages will have died out, and some other lineages will have acquired new young branches. Then in the third generation the same thing happens, and more old lineages are lost. Eventually you will reach a generation when everyone in the population is descended from just one of the original men. This depends on chance but on average takes as many generations as the effective male population size. So after 500 years, say, that will probably be the case - one lineage may become fixed, and the TMRCA of the lineage in that population will be 500 years.

Now after another 500 years, one of the sublineages carried by the 20 men who lived 500 years after the founding has become fixed. So the TMRCA of the lineage carried by the population is still 500 years. But if you look at the tree of that lineage, you will see 500 years of branching, and before that 500 years of *no* branching in the lineage. Not because the lineage didn't branch - we know it did - but because only one branch survived.

And that is why we expect to get those long sequences of SNPs. It is perfectly normal and there is no need to invoke processes causing runs of mutations. Which is not to say such processes couldn't exist...

George Chandler
09-12-2015, 06:03 PM
It's not meant to wear you out miser and I'm not saying you're wrong only that what you're suggesting isn't (IMO) likely the only reason for what you're seeing and I'm definitely not convinced yet that it's the main cause.


One might suppose that these 20 males are all going to be from the same branch after such a long bottleneck, but this couldn't be further from the truth.

Not necessarily. Even though a family could be part of a larger tribal community or area they may geographically isolated. So let's say you have a new SNP within a new family and one son heads for the mountains and starts his family there and becomes a somewhat isolated tribe over a thousand years. His small tribe of descendants suffer events such as plagues, starvation etc and is left with very few males. It doesn't have to be the last man or only a single man all it has to be is the number of men in that family group between that last SNP they share. His bothers stay in the lowlands and over the same thousand years become different small communities. An invading army enters the lowlands and wipes out the tribes to the last man. Another example is a that small family group from that lowland community (say 20 males of 3 generations) who all share a most recent SNP and no other mutation has happened since. They were off trading somewhere with another tribe and were spared the destruction by the invading army. It's likely you wouldn't be able to distinguish using SNP's if it was a single man left or all 20 other than maybe the rate of expansion after the genetic bottleneck? You would be able to see the difference between the mountain community and the surviving lowland family because they were separated by a thousand years.


So even if many extended duration bottlenecks have occurred in all the major haplogroups, as some people suppose they have, they would not have created long stringy lineages without branching, with numerous equivalent SNPs.

Not necessarily true. It depends on how many similar events took place within that culture. It could be some genetic abnormality that the ancestors passed down in terms of a disease shortening the life span? You can have that 20 people which are left who's male population expands rapidly and have another similar event (such as the invasion example) happens thousands of years later where the population is knocked down to a dozen small close family groups who's male population then starts expanding again. It doesn't have to be reduced to a single man.

George

JRW
09-12-2015, 06:55 PM
Those of you engaged in this debate should take a look at the following paper regarding calculating TMRCA ages if you have not done so already. It is very germane to this discussion: http://biorxiv.org/content/early/2015/09/08/026286.full-text.pdf+html.

miiser
09-12-2015, 08:14 PM
Those of you engaged in this debate should take a look at the following paper regarding calculating TMRCA ages if you have not done so already. It is very germane to this discussion: http://biorxiv.org/content/early/2015/09/08/026286.full-text.pdf+html.

Thanks for sharing this. Unfortunately, this paper is applicable to STR mutations, but not generally to SNP mutations.

"Our main ideas hold for general molecular clocks, however our Y-clock has features specific to STR mutations."

SNPs and STRs occur through fundamentally different processes, and do not follow the same mathematical rules. SNP count has a greater branch to branch variation than STRs. My concerns apply to SNP counting.

miiser
09-12-2015, 08:17 PM
New branches form and old branches die out. Say you have a small population founded by 20 men, representing 20 lineages. In each generation some of those men will have no children, some will have only daughters, some will have just one son, some will have many sons. In the next generation, if the population does not grow, there will still be the same number of men, but now some of the original lineages will have died out, and some other lineages will have acquired new young branches. Then in the third generation the same thing happens, and more old lineages are lost. Eventually you will reach a generation when everyone in the population is descended from just one of the original men. This depends on chance but on average takes as many generations as the effective male population size. So after 500 years, say, that will probably be the case - one lineage may become fixed, and the TMRCA of the lineage in that population will be 500 years.

Now after another 500 years, one of the sublineages carried by the 20 men who lived 500 years after the founding has become fixed. So the TMRCA of the lineage carried by the population is still 500 years. But if you look at the tree of that lineage, you will see 500 years of branching, and before that 500 years of *no* branching in the lineage. Not because the lineage didn't branch - we know it did - but because only one branch survived.

And that is why we expect to get those long sequences of SNPs. It is perfectly normal and there is no need to invoke processes causing runs of mutations. Which is not to say such processes couldn't exist...

I expected that someone would point out that certain Y lineages dominate over time and others go extinct. But this is a different process than the one that I pointed out. Even if a lineage dominates a population over a long time period, you will still have haplogroup divisions in the tree BELOW the MRCA. The descendants that eventually lead to person A's lineage dominating will not all belong to the same descendant branch of person A. There will be different descendant branches with different SNPs. And so this will still not create a large block of phylogenetically equivalent SNPs as the result of a population bottleneck.

Edit for clarification of my stance: I went too far in my previous post, saying that equivalent blocks could ONLY be caused by rapid runs. You are correct that they can also be created by dominance of a single Y lineage. However, my main point still stands. They aren't created by population bottlenecks. This is the point I was making with the 20 member population example. Equivalent blocks are not a reliable indicator of bottlenecks. And even if lineage dominance is able to explain some of the equivalent blocks, we still have to address the large number of branches per node, the concentration of the total number of branches within a surprisingly small number of nodes, and the large branch to branch variation of SNP count. And I believe these can only be accounted for by SNP runs.

JRW
09-13-2015, 04:59 AM
SNPs and STRs occur through fundamentally different processes, and do not follow the same mathematical rules. SNP count has a greater branch to branch variation than STRs.

I am not sure I follow your comment. By definition mathematical models based on STR variables rather than SNP variables to describe the underlying phylogenetics would be different. However, in either case both types of approaches are modeling the evolution of the same human lineages. As you will note in the paper, the STR-based model developed by the author is used to calculate the ages of SNP-defined lineages.

miiser
09-13-2015, 05:50 AM
I am not sure I follow your comment. By definition mathematical models based on STR variables rather than SNP variables to describe the underlying phylogenetics would be different. However, in either case both types of approaches are modeling the evolution of the same human lineages. As you will note in the paper, the STR-based model developed by the author is used to calculate the ages of SNP-defined lineages.

Your first comment, posting the link to the publication, suggested that it is relevant to the debate in this thread. I agree with the mathematical approach used and conclusions reached in the publication. However, the current debate in this thread has to do with using SNP counting, not STR mutations, to make certain types of deductions. Specifically, the current debate deals with the uneven distribution of SNPs. This is a detailed aspect of SNP based molecular clocks which is not addressed in the publication. And so the publication is not particularly applicable to this current debate.

The debate revolves around whether the uneven distribution of SNPs impairs the ability to use SNP counting to infer population bottlenecks or haplogroup growth rate, and limits the ability to accurately estimate the date of individual nodes along a branch. SNPs and STRs occur by a different molecular process. The publication claims to prove that STRs occur at a reasonably constant rate in order to enable certain dating techniques to work when adjustments are made for non-molecular effects such as longevity, reproductive advantage, etc. The publication does NOT similarly show that SNPs occur at a reasonably constant rate. If SNPs are less evenly distributed, then this impairs the ability to deduce haplogroup growth rate based on the number of branches or number of phylogenetically equivalent SNPs at a node. It also increases the error range of SNP count based age estimates and significantly impairs the ability to estimate the timing of specific nodes along a branch.

The current debate revolves around the validity of a particular argument based on SNP counting. Never in this debate have I argued that M222, or any other haplogroup, is or isn't a particular age, or that it did or did not grow at a certain pace in a certain time period. This debate was started because an SNP counting method was incorrectly applied to deduce a population bottleneck and subsequent growth spurt. I don't claim to know when M222 had bottlenecks or growth spurts. I only argue that phylogenetically equivalent blocks of SNPs and number of branches per node can't be used to support such claims.

I agree that the publication is interesting and provides an improved model for STR based dating. I just don't think it's particularly relevant to the current debate.

Maybe you could help me out by explaining why, specifically, you think it's germane to the current debate.

Megalophias
09-13-2015, 06:40 AM
Equivalent blocks are not a reliable indicator of bottlenecks. And even if lineage dominance is able to explain some of the equivalent blocks, we still have to address the large number of branches per node, the concentration of the total number of branches within a surprisingly small number of nodes, and the large branch to branch variation of SNP count. And I believe these can only be accounted for by SNP runs.
Have you actually worked this out mathematically, or is this just an intuition?

miiser
09-13-2015, 07:33 AM
Have you actually worked this out mathematically, or is this just an intuition?

Fairly simplistic models and mathematics, yes. I haven't gone to the trouble to code a robust software model.

Fairly early after Big Y data first became available in significant data size, I thought the number of new branches being discovered was smaller than it ought to be and the large number of branches per node was greater than it ought to be (intuition). So I made some fairly simple models using Excel spreadsheets to investigate (math). Using an SNP rate of 1 per 3 generations (in the early days, the average mutation rate was thought to be closer to 1 per 90 years), I was not able to model a tree with such large numbers of branches per node to match what we observe in M222 and throughout L21. I was only able to reproduce it by significantly increasing the percentage of offspring containing no new SNPs well beyond 3, 4, or 5 generations between SNPs. Assuming the accepted average mutation rate is close to correct, this necessitates collapsing some of the single SNPs into SNP runs, which effectively increases the time between branch creating SNPs. I also spent some time measuring the branch to branch variation of SNP count, and confirmed that it was greater than it should be in the case of single, random SNPs.

Now people are going to ask me to upload my work or provide tables, images, or specific numbers for my calculations. And I will give the lame answer that the work was done some time ago, I can't easily find it now (having gotten a new computer less than a year ago), and the presentation was very messy so that it will not be comprehensible to anyone other than myself unless I put a large amount of effort into basically redoing it from scratch and cleaning it up to make it presentable. I don't have the time to redo and polish the work now and am unwilling to do it, having other things of higher priority in my life. So people can, in fairness, say that I should provide hard evidence to back up such claims. (And I would counter that they should provide hard evidence to demonstrate that SNPs occur at a uniform rate.)

But, hey, if someone else can argue, "I've done a similar study, and my work shows that the tree structure matches single, random SNPs," then I'd be willing to consider their evidence or arguments and ready to consider that my models were wrong. The argument I'm not ready to accept is when people say or imply, "No, you just don't understand how branch formation works, and I've been a well respected poster on this forum for way longer than you, so I dismiss your argument out of hand," . . . or something substantially equivalent to this. (You haven't done this to me, so don't interpret this comment as being directed at you.) And I think it's poor science for people to continue interpreting SNP counts under the ASSUMPTION that SNPs ARE evenly distributed, in spite of the fact that there have been several people arguing that there is significant evidence to the contrary. And no one, as far as I'm aware, has provided any evidence to demonstrate that SNPs ARE evenly distributed as single, isolated events.

JRW
09-13-2015, 01:22 PM
Maybe you could help me out by explaining why, specifically, you think it's germane to the current debate.

The thrust of the paper is about developing an accurate algorithm to model phylogenetic branching and ultimately, TMRCA age calculations. The author reviews and cites previously developed models -- some SNP based -- in the paper. Isn't that the core topic of this debate? IMO, if the topic is only as narrow as SNP counting, then we're wasting our time. The paper clearly outlines, even to a layman, that the counting of mutational events (don't get fixated on the type of mutational event -- that is the gnat in a room of elephants) alone does not sufficiently model phylogenetic branching.

miiser
09-13-2015, 09:51 PM
The thrust of the paper is about developing an accurate algorithm to model phylogenetic branching and ultimately, TMRCA age calculations. The author reviews and cites previously developed models -- some SNP based -- in the paper. Isn't that the core topic of this debate? IMO, if the topic is only as narrow as SNP counting, then we're wasting our time. The paper clearly outlines, even to a layman, that the counting of mutational events (don't get fixated on the type of mutational event -- that is the gnat in a room of elephants) alone does not sufficiently model phylogenetic branching.

I don't think anyone in this debate is arguing that SNP count is all that matters. But if an SNP counting method is being widely applied by numerous people in the forum to make deductions, and is assumed by many forum readers to be a valid method, then it's important to establish whether or not this method is valid . . . and if it is NOT valid, it's important to convince people to stop using it and stop trusting it. I can't see that this is "wasting our time".

There are a variety of methods and tools available to date haplogroups. All of the methods that are valid should be applied in parallel to provide verification. The fact that there are multiple available methodologies available does not negate the need to prove that each one of them is valid. This debate is useful because the goal is to correct false assumptions and establish a consensus regarding whether or not one particular method is valid.

George Chandler
09-13-2015, 10:30 PM
(And I would counter that they should provide hard evidence to demonstrate that SNPs occur at a uniform rate.)


Unless I've missed something from an earlier post I don't think anyone is saying that SNP's occur at a uniform rate..they are a random event. This is why I'm saying that you're correct that there is the "possibility" of runs or gaps in the number of years per mutation. It doesn't mean it's likely as mentioned previously. Determining the number of years on average between each mutation or each Sanger validated mutation is an average based on comparing validated ancestral lines and using proven genealogy. Take that and combine DNA and carbon dating analysis from human remains and it gets you close. It's not an absolute and nothing is unless you are targeting a specific ancient person who's remains you can test and validate.

It's hard to understand why you're pushing your analysis but can't support it?

miiser
09-14-2015, 12:05 AM
Unless I've missed something from an earlier post I don't think anyone is saying that SNP's occur at a uniform rate..

People are assuming implicitly, without stating explicitly, that SNPs occur at a uniform rate with comments such as "The 42 phylogenetically equivalent SNPs show that M222 had a very long population bottleneck." That you don't see the inherent assumption in such statements only serves to convince me that you don't understand how SNP count based dating works.


they are a random event. This is why I'm saying that you're correct that there is the "possibility" of runs or gaps in the number of years per mutation. It doesn't mean it's likely as mentioned previously.

This comment demonstrates that you do not even understand the argument I am making. SNPs are NOT random, unrelated events. The rate depends on physical processes which are affected by various factors, such as age of the father, environmental effects, and interactions with other SNPs. Because of this, they are not randomly distributed, but are sometimes (often enough to be statistically significant to SNP counting based methods) temporally grouped into clumps.


Determining the number of years on average between each mutation or each Sanger validated mutation is an average based on comparing validated ancestral lines and using proven genealogy. Take that and combine DNA and carbon dating analysis from human remains and it gets you close. It's not an absolute and nothing is unless you are targeting a specific ancient person who's remains you can test and validate.

It's hard to understand why you're pushing your analysis but can't support it?

I have stated several times very clearly that the large number of branches per node and large branch to branch variation supports my claim. But, rather than address this argument directly, your responses have repeatedly diverted the conversation into irrelevant side topics such as sampling rate, sequencing methods, Y lineage dominance, etc. under the false assumption that these factors are capable of generating the observed tree structure. As far as I can tell, Megalophias and I are the only two people in this discussion who even understand the topic well enough to have an intelligent dialog. For this reason, I will continue to hold debate with him or others with rational arguments. But I choose to no longer debate with someone who doesn't even address the arguments I'm making. I refuse to humor you by pretending that the unrelated topics you keep bringing into the discussion are capable of generating the observed tree structures, when I've gone through the trouble of modelling the tree and know that they aren't.

Iridescent calcium carbonate concretions and suidae - the reason all the smartest people eventually give up and leave this forum. Consider this comment as goodbye and farewell. Future non argument responses from you will be ignored.

George Chandler
09-14-2015, 01:15 AM
Unless a haplogroup experiences a bottleneck which cuts it down to literally just a SINGLE individual, there will still be multiple branches containing different SNPs, even in a very small population. And hypothetical bottlenecks which constrain a haplogroup population to just a single individual without extinguishing it entirely are so remotely improbable that they are not even worth considering as a reasonable explanation for an SNP block.


This coming from a person who wrote this incorrect statement. Maybe you should go to another forum where your high intellectual abilities can be seen for what they are and they can stroke you for being so wise in the way of genetics.

Have fun with your research!!:)

jdean
09-14-2015, 11:12 AM
SNPs are NOT random

So this is just a continuation of your theory that STR mutations aren't random which you've been posting on various forums for years.

GogMagog
09-15-2015, 03:36 PM
Play nice.

AtWhatCost
09-17-2015, 07:43 PM
I actually don't even think M222 is connected to Niall, I think he was a different snp, call it women's intuition, I don't have any spread sheets or anything to support it, but Niall wasn't M222.

Dubhthach
09-17-2015, 08:06 PM
It's debatable if Níall even existed tbh

miiser
09-17-2015, 10:39 PM
So this is just a continuation of your theory that STR mutations aren't random which you've been posting on various forums for years.

I have NEVER argued that STRs exhibit the same uneven distribution as SNPs, on this forum or on any other. On the contrary, my previous comments in this thread have made it clear that I believe there are real differences in the way STRs mutate versus SNPs. SNPs exhibit a larger variance than do STRs. STR mutations are distributed in a way that is mostly consistent with randomness. SNPs are distributed in a way that is NOT consistent with randomness. The test of randomness in statistics is to compare the distribution to a random binomial distribution. The standard deviation of observed SNPs is much greater than the standard deviation of a binomial distribution. This shows that SNPs fail the binomial test and are, therefore, probably not randomly distributed.

I would prefer if you did not attribute straw man arguments to me which I have not made. I believe this is a violation of forum rules. If you believe I've made such an argument, please quote it here.

jdean
09-17-2015, 11:38 PM
I have NEVER argued that STRs exhibit the same uneven distribution as SNPs, on this forum or on any other. On the contrary, my previous comments in this thread have made it clear that I believe there are real differences in the way STRs mutate versus SNPs. SNPs exhibit a larger variance than do STRs. STRs are distributed in a way that is mostly consistent with randomness. SNPs are distributed in a way that is NOT consistent with randomness. The test of randomness in statistics is to compare the distribution to a random binomial distribution. The standard deviation of observed SNPs is greater than the standard deviation of a binomial distribution. This shows that SNPs fail the binomial test and are, therefore, probably not randomly distributed.

I would prefer if you did not attribute straw man arguments to me which I have not made. I believe this is a violation of forum rules. If you believe I've made such an argument, please quote it here.

If you aren't the person who used to trade as Ironroad41 & Mcg11 then I've spoken out of order, however there are a lot of similarities in your theme.

George Chandler
09-18-2015, 12:34 AM
If you aren't the person who used to trade as Ironroad41 & Mcg11 then I've spoken out of order, however there are a lot of similarities in your theme.

In one statement he says the following:


I have stated several times very clearly that the large number of branches per node and large branch to branch variation supports my claim. But, rather than address this argument directly, your responses have repeatedly diverted the conversation into irrelevant side topics such as sampling rate, sequencing methods, Y lineage dominance, etc. under the false assumption that these factors are capable of generating the observed tree structure.

He then goes on to state that SNP's are not consistent with randomness. I'm currently doing a spreadsheet and it shows the differences between Sanger validated and non Sanger validated SNP positions. Some (being ancestral) are popping up intermittently with some people being positive and some peoples test kits being negative "BECAUSE THEY ARE UNSTABLE". The Sanger validated ones are right across all test kits and positive (where they should be) unless there was a missing SNP from a sequencing error (which has also happened). Could some of these non Sanger validated SNP's be valuable positions?...absolutely, but the very fact that they are "randomly" showing up in some results and not others (yet ancestral to all) would prove randomness exists right there. Of course though..his "Big Y research" shows that SNP's are not consistent with "randomness".

If he wants to find randomness in M222 all he has to do is validate through the Sanger method and determine which are ancestral then compare all kits and plot the ones which don't meet Sanger validation. What about the positions which are found in two different kits yet are two totally different mutation events of the same position?

Do waste your time on his argument.

George

miiser
09-18-2015, 02:34 AM
If you aren't the person who used to trade as Ironroad41 & Mcg11 then I've spoken out of order, however there are a lot of similarities in your theme.

I have only one account on Anthrogenica and rarely post on any other public forum. There are very few similarities between my posts and those other posters. I have sometimes argued against mcg11 in past threads. Whatever similarities you may think you see between us, the differences are far greater. I use rational arguments to make my points. They, typically, do not. I am not attempting to support or refute any particular viewpoint regarding the origin or age of a particular group. I am only challenging the validity of arguments that are based on unjustifiable assumptions regarding the nature of SNPs. The outcome of the data analysis - whether the M222 population was bottle necked or growing quickly during a particular time - is a piece of trivia that is not of much interest to me. I have no personal stake in the age, origin, and history of M222, or of any other haplogroup. My only agenda is to correct a community misstep in the way SNP counts are being interpreted by many and to encourage the proper application of SNP counting.

jdean
09-18-2015, 10:32 AM
I have only one account on Anthrogenica and rarely post on any other public forum. There are very few similarities between my posts and those other posters. I have sometimes argued against mcg11 in past threads. Whatever similarities you may think you see between us, the differences are far greater. I use rational arguments to make my points. They, typically, do not. I am not attempting to support or refute any particular viewpoint regarding the origin or age of a particular group. I am only challenging the validity of arguments that are based on unjustifiable assumptions regarding the nature of SNPs. The outcome of the data analysis - whether the M222 population was bottle necked or growing quickly during a particular time - is a piece of trivia that is not of much interest to me. I have no personal stake in the age, origin, and history of M222, or of any other haplogroup. My only agenda is to correct a community misstep in the way SNP counts are being interpreted by many and to encourage the proper application of SNP counting.

I completely fail to see how rapid expansion of a group (which presumably did happen with the M222 block) would generate a large number of equivalent SNPs.

Dubhthach
09-18-2015, 11:20 AM
Meanwhile back in M222 land I'm awaiting on a A259 result for a Connor who has matches with other O'Connor's/McManus with background in Connacht, he's also ordered upgrade to 67 STR's and the M222 bundle.

It will be interesting to see if he comes back as A259+/A260- like some of his McManus matches. The hint here is that McManus of Connacht family are a branch of O'Connor's of Connacht with seperation point been in late 12th/early 13th centuries.

MichaelK
10-02-2015, 12:13 PM
Alot of information/data for a noob. Looking forward to reading the discussions.

(M222+DF109/S660+S588-A223+pending)
23andme, FtDNA, an YSEQ(3331) testing

dp
11-05-2015, 03:45 PM
Alot of information/data for a noob. Looking forward to reading the discussions.

(M222+DF109/S660+S588-A223+pending)
23andme, FtDNA, an YSEQ(3331) testing

A belated Welcome to Anthrogenica :-)
-dp

Dubhthach
11-20-2015, 10:04 AM
Meanwhile back in M222 land I'm awaiting on a A259 result for a Connor who has matches with other O'Connor's/McManus with background in Connacht, he's also ordered upgrade to 67 STR's and the M222 bundle.

It will be interesting to see if he comes back as A259+/A260- like some of his McManus matches. The hint here is that McManus of Connacht family are a branch of O'Connor's of Connacht with seperation point been in late 12th/early 13th centuries.

This "Connor" kit is now confirmed as A259+ (snp test) and A260- (M222 bundle), at 67 markers he shows matches with both McManus kits. Given that vast majority of A259 so far is also A260+, it's interesting to see A259+/A260- bearing surnames linked to Roscommon/North Connacht, it might hint at point of origin of clade within wider DF105.

Rory Cain
11-20-2015, 11:25 AM
This "Connor" kit is now confirmed as A259+ (snp test) and A260- (M222 bundle), at 67 markers he shows matches with both McManus kits. Given that vast majority of A259 so far is also A260+, it's interesting to see A259+/A260- bearing surnames linked to Roscommon/North Connacht, it might hint at point of origin of clade within wider DF105.

Good. May now we should be seeing where the Ui Fiachra, Ui Briuin, Ui Niall etc branched off. We may even be on the verge of where individual clans branched off- O'Connor, O'Donnell, O'Cathain, O'Neill, etc.

Dubhthach
11-20-2015, 12:23 PM
The Uí Fiachrach type surnames are generally undifferenated DF105, though interesting Alex Williamson and co have idendentifed a SNP called FGC23742 which is eating into the undifferenated DF105.

We've got at least one O'Donnell who is DF85+, along with several Doherty's, Boyle's, Gallaghers etc. S588 is other major northern subclade of DF105. Both DF85 and S588 make showing among scottish men as well, in comparison A259 is nearly entirely dominated by men with Uí Briúin type surnames.

Rory Cain
11-20-2015, 08:25 PM
The Uí Fiachrach type surnames are generally undifferenated DF105, though interesting Alex Williamson and co have idendentifed a SNP called FGC23742 which is eating into the undifferenated DF105.

We've got at least one O'Donnell who is DF85+, along with several Doherty's, Boyle's, Gallaghers etc. S588 is other major northern subclade of DF105. Both DF85 and S588 make showing among scottish men as well, in comparison A259 is nearly entirely dominated by men with Uí Briúin type surnames.

The O'Cathains of Ulster were one of the few groups to gain anything from the original Deep Clade test, which was too Basque-centric for Isles folks. Having gained M222, they have largely ignored NGS. Conversely the O'Cathains of Galway having gained nothing from Deep Clade have pushed on with NGS, only to find that their Ui Fiachra genealogy must have been a grant for services, like a knighthood, rather than a factual pedigree. They are R-DF21 like their followers, the O'Mochain who were also granted an Uo Fiachra pedigree.

Finally we have some SNP results from O'Cathains of Ulster who are M222 > DF105 > S589 > BY202 along with, as you said, mostly Scots surnames. But in general we need more NGS results from this group and are not getting them, as most seem to be comfortable at having gained M222 from the old Deep Clade test that failed almost everyone else.

Rory Cain
11-20-2015, 08:31 PM
... A259 is nearly entirely dominated by men with Uí Briúin type surnames.

O'Shaughnessy is an Ui Fiacha Aidhne surname. He is A259.

AtWhatCost
11-21-2015, 04:32 PM
O'Shaughnessy is an Ui Fiacha Aidhne surname. He is A259.

How large is the sample size for the surname, I only see 4 members in that surname project. Do we have enough samples to even propose a connection of any sort? What are the 4 members that are currently in the Shaughnessy project, as far as predicted snps. Are there are a significant number of this surname somewhere else tested. I tend to see a lot of cherry picking singletons and proposing connections.

dp
11-21-2015, 05:31 PM
How large is the sample size for the surname, I only see 4 members in that surname project. Do we have enough samples to even propose a connection of any sort? What are the 4 members that are currently in the Shaughnessy project, as far as predicted snps. Are there are a significant number of this surname somewhere else tested. I tend to see a lot of cherry picking singletons and proposing connections.
The sharing of SNPs is a criteria in genetic genealogy. If we assume the progenitor of a family lived at the time surnames came into vogue, say 500 years ago --if not 600, the rate that an SNP would become fixed should be sufficient to define branches within that family. If people of a particular surname, with a large enough genetic difference between STR results, share a phylogenetic SNP then it is reasonable that it is an indicator of a common ancestor. This doesn't make the initial SNP carrier have to have had that surname though; but until people that have other surnames are found that have that SNP, it is a reasonable indicator that either it belonged to an ancestor of the founder of the surname in question, and that all brother (mutual SNP carrier) branches have died out (or are untested), or that the surname founder indeed had the SNP mutation, or that a person bearing that surname had the SNP and passed it to their sons.
Yours,
dp
PS: haven't had lunch so might have something off

Dubhthach
11-21-2015, 05:53 PM
How large is the sample size for the surname, I only see 4 members in that surname project. Do we have enough samples to even propose a connection of any sort? What are the 4 members that are currently in the Shaughnessy project, as far as predicted snps. Are there are a significant number of this surname somewhere else tested. I tend to see a lot of cherry picking singletons and proposing connections.

The specific O'Shaughnessy is A259+/A260+ (McManus and the specific Connor kit are both A259+/A260-) interesting enough in Ireland project we have another O'Shaughnessy tested to 67 markers and is M222+. The two of them have GD of 13 at 67 markers!

In Genetic Distance report the A260+ O'Shaughnessy unsurprising matches closest to other A260+ men, most of whom bear more "northern Connacht" surnames than would be expected for a Uí Fiachrach Aidhne family (eg. South Galway, O'Shaughnessy lordship was centered on Gort until the 17th century.

When we look at the other O'Shaugnessy (who hasn't done M222 bundle) interesting at 67 markers he has a match also bearing surname Shaughnessy (GD of 5), he has another Shaugnessy match that shows up at 25 STR's. Without this kit doing BigY or at least the M222 bundle we won't know if there's connection or not.

As for the surname, interesting enough Shaughnessy is one of those rare Irish surnames that has only one documented occurence, without testing the other 3 or so O'Shaughnessy's that I can see for A259 we won't know if it's relevant to family or not.

Rory Cain
11-21-2015, 09:15 PM
I tend to see a lot of cherry picking singletons and proposing connections.]

A lot of singleton's ? Truly? By definition a singleton would be one. Yes, O'Shaughnessy is a singleton (as far as the Ui Fiachra Aidhne go). And yes, he is A259 unless you believe the lab got that wrong. Given the the Ui Briuin test results are tending to be A259, and Fiachra was the brother of Briuin, do you have grounds to discard this otherwise plausible result?

Dubhthach
11-22-2015, 09:47 AM
As noted above I'm wary of the O'Shaughnessy A259+/A260+ result. I think we need to at least test the other three O'Shaughnessy's to do following:
1. If M222 (or predicted) are they A259+ or A259-
2. If A259- than test for DF105

Of course the M222 bundle is available for $109 at moment, I believe (but could be wrong) that A259 was added to "version 2" of this bundle, unfortunately it wasn't included in first run (A260 was)

Rory Cain
11-22-2015, 11:29 AM
As noted above I'm wary of the O'Shaughnessy A259+/A260+ result. I think we need to at least test the other three O'Shaughnessy's ...

Agreed. One O'Shaughnessy R-A260 alone is not enough. But nor should his result be discarded, as another correspondent suggests. While O'Shaughnessy is a sept of Ui Fiachra Aidhne of South Connaught, let us not forget that the Ui Fiachra Muaidhe of North Connaught share the same descent. That could explain O'Shaughnessy's north Connaught matches. These include persons named Conley, Connolly, Curran, Donohoe, Flynn and McHugh who are also R-A260. These surnames, some of which admittedly have multiple origins, are associated with the Ui Fiachra Muaidhe of North Connaught.

Plus three Ui Fiachra Aidhne clans have a DNA Project: O'Shaughnessy, as previously noted, with 14 members, of whom 4 have made their results public, so who know what to make if them; but also O'Cleirigh, who have 36712 O'Shaughnessy grouped with a Cleary and two Clarkes (the anglicised used form of O'Cleirigh); and the Cahill DNA Project, who have a R-A260 member grouped with other R-M269 Cahills who may well prove to be R-A260. Three known Ui Fiachra Aidhne surnames plus five potentially Ui Fiachra Muaidhe surnames is more than a reasonable person would discard.

Far too early to shut the door on O'Shaughnessy. The R-L1402 group also included a mix of surnames which confused people until I pointed out that it also includes five of The Seven Septs of Laois, a branch of the Clanna Rory. Likewise R-L1336 includes a mix of surnames, amongst which are nearly hidden the surnames of the Corcomroe branch of the Clanna Rory.

One only sees 10% of the iceberg, so I'm not making a firm connection yet, without seeing the rest of the iceberg. But it would seem rash to discard what we see thus far.

AtWhatCost
11-22-2015, 11:33 PM
]

A lot of singleton's ? Truly? By definition a singleton would be one. Yes, O'Shaughnessy is a singleton (as far as the Ui Fiachra Aidhne go). And yes, he is A259 unless you believe the lab got that wrong. Given the the Ui Briuin test results are tending to be A259, and Fiachra was the brother of Briuin, do you have grounds to discard this otherwise plausible result?

Where do you see the words a lot of singletons? The words typed were, a lot of cherry picking. Next time, read a little slower, it tends to help.

AtWhatCost
11-22-2015, 11:35 PM
]

A lot of singleton's ? Truly? By definition a singleton would be one. Yes, O'Shaughnessy is a singleton (as far as the Ui Fiachra Aidhne go). And yes, he is A259 unless you believe the lab got that wrong. Given the the Ui Briuin test results are tending to be A259, and Fiachra was the brother of Briuin, do you have grounds to discard this otherwise plausible result?

It's a singleton, no need too claim the clan has been found, a bit premature I think.

AtWhatCost
11-22-2015, 11:36 PM
Until NIall has been found, and his DNA is analyzed, I'm also not buying he was M222.

Rory Cain
11-23-2015, 01:36 AM
Where do you see the words a lot of singletons? The words typed were, a lot of cherry picking. Next time, read a little slower, it tends to help.

How you believe Alex cherry-picked the results in his Big Tree is a figment of a fevered imagination. Customers select themselves for testing. Some, perhaps many, then choose to share those results with Alex. I know Alex better than to believe he is doing any cherry-picking.

Dubhthach
11-23-2015, 07:16 AM
It's just as likely that Niall never existed ;) Likewise for his three half brothers. After all that's what we are talking about not the Uí Néill but the Connachta who in genealogical tradition do not descend from the bould Niall but from his supposed half-brothers.

What's more interesting is that the genealogical structure of medieval genealogies seem to have some basis in fact when we look at the Uí Briúin anyways. In the historical narrative the Uí Briúin only come to prominence around 690-700AD when they (in the form of the Uí Briúin Aí take the Kingship of Connacht from Uí Fiachrach for the first time).

So far we have A259+ tested men with surnames belonging to Uí Briúin Aí, Uí Briúin Bréifne and Uí Briúin Seola, within the genealogical framework of medieval Ireland these three kindreds seperated from each other during the 6th century. Obviously the earliest strata of the genealogy is probably falsified, but given common SNP it could be case that they were woven together (in perhaps a fanciful form) to show that within context of wider Dál Cuinn/Connachta that they were more closely connected to each other.

I should note again that a close male relative of the O'Conor Don (who has a unbroken genealogy as head of the O'Connor family of Connacht -- recognised by both the British and Irish states) is M222+, though he hasn't tested for A259 he has close matches with a number of A259+/A260- men.

AtWhatCost
11-23-2015, 03:09 PM
How you believe Alex cherry-picked the results in his Big Tree is a figment of a fevered imagination. Customers select themselves for testing. Some, perhaps many, then choose to share those results with Alex. I know Alex better than to believe he is doing any cherry-picking.

Once again, I don't see any post where I even mentioned the name of Alex or referred to his tree. Could you please stick with what people actually type and not twist information, respectfully? Thank you for the personal attack though on my "fevered imagination". I will be a man and ignore that particular comment. Now, if we can get more than 1 or 2 kits before we pin the name to a clan, that would be great. Also, I'm waiting for Niall's remains to confirm that M222 status also. Thank you.

Dubhthach
11-23-2015, 03:40 PM
I should note that "Clann" in Irish doesn't mean family of family group, it literally means ones children. That and it's utilatmely a loan word from Latin via Old Welsh:

Planta (Latin) -> Plant (Welsh) -> Cland (Old Irish) -> Clann (Modern Irish) -> Clan (English)

Again as mention it's just as likely that Niall didn't actually exist, he's on the twighlight between pre-christian and christian era's. As a result he might be a literary construct (the main story about him is high saga), or as some historians note he might have been a rather minor character whose main claim to fame was his son's.

An interesting foreign comparison to him is to look at Ragnar Lodbrok (Ragnarr Loðbrók), who apart from been the main character on "Vikings" tv show is supposed father of multiple historic kings in Scandinavia. He himself possibly didn't exist.

Anyways if my theory (emphasis on theory) that A259 is linked to various Uí Briúin dynastical groups of Medieval Connacht is correct, than whether Niall was M222+ or not is immaterial. The Uí Briúin like the other 2 Connachta (Uí Fiachrach -- thence talk about Shaughnessy) and Uí nAilleo (wiped out in 8th century by expansion of Uí Briúin) aren't descended from Niall anyways, but from 3 of his supposed half-brothers eg. Brion, Fiachrach and Aillil.

AtWhatCost
11-23-2015, 04:23 PM
I should note that "Clann" in Irish doesn't mean family of family group, it literally means ones children. That and it's utilatmely a loan word from Latin via Old Welsh:

Planta (Latin) -> Plant (Welsh) -> Cland (Old Irish) -> Clann (Modern Irish) -> Clan (English)

Again as mention it's just as likely that Niall didn't actually exist, he's on the twighlight between pre-christian and christian era's. As a result he might be a literary construct (the main story about him is high saga), or as some historians note he might have been a rather minor character whose main claim to fame was his son's.

An interesting foreign comparison to him is to look at Ragnar Lodbrok (Ragnarr Loðbrók), who apart from been the main character on "Vikings" tv show is supposed father of multiple historic kings in Scandinavia. He himself possibly didn't exist.

Anyways if my theory (emphasis on theory) that A259 is linked to various Uí Briúin dynastical groups of Medieval Connacht is correct, than whether Niall was M222+ or not is immaterial. The Uí Briúin like the other 2 Connachta (Uí Fiachrach -- thence talk about Shaughnessy) and Uí nAilleo (wiped out in 8th century by expansion of Uí Briúin) aren't descended from Niall anyways, but from 3 of his supposed half-brothers eg. Brion, Fiachrach and Aillil.

Thank you for the thoughts. We (the community) don't have enough results to even determine the main Shaughnessy lineage is A259 or M222 for that matter (again, very limited testing), we only know of the small number that are. So I agree, Niall is moot (I always thought he was anyway..) but there still isn't anything other than a very small number of results to connect Shaughnessy to this group. I for one believe there is a good chance the Ui Fiachrach are a completely different snp unrelated to M222 but, there is nowhere near significant evidence to substantiate what snp that might be, not at this time. So to circle back around, I personally don't feel there is nearly enough evidence to demonstrate with any high degree of certainty, the main Shaughnessy lineage of the Ui Fiachrach is A259, that's pretty much it and I don't think anything related to this can be proven otherwise at this time. Even if Niall existed, since everything appears to be way over-blown I wouldn't be the least bit surprised if he only had one or two hostages at best :)

Edit Info: My personal interest is who/what groups were in the area prior to the invasions of the Connachta and the Ui Neill and what are their SNPs.

Dubhthach
11-23-2015, 04:38 PM
I understand however, we don't have enough results to even determine the main Shaughnessy lineage is A259, we only know of the small number that are. So I agree, Niall is moot then but there still isn't anything other than a very small number of results to connect Shaughnessy to this group. I for one believe there is a good chance the Ui Fiachrach are a completely different snp unrelated to M222 but, there is nowhere near significant evidence to substantiate what snp that might be, not at this time. So to circle back around, I personally don't feel there is nearly enough evidence to demonstrate with any high degree of certainty, the main Shaughnessy lineage of the Ui Fiachrach is A259, that's pretty much it.

As I've pointed out in previous posts this one A259+/A260+ Shaughnessy is at quite a genetic distance from other M222+ O'Shaughnessy's. My own feeling is that other ones if tested would be A259-, this is in part driven by observation that men with Uí Fiachrach type surnames who've done BigY/M222 bundle are often DF105* (eg. negative for all subclades of DF105 such as A259, S588, DF85 etc.) -- though this is complicated by fact that A259 wasn't included in first M222 bundle (A260 was include -- they just launched revised bundle). A number of these men who have tested individually for A259 have come back negative.

Thence my feeling that A259 is specifically the Uí Briúin lineage. Interesting enough there is a new SNP on the block under DF105 which might make significant inroads into the DF105* grouping. This is: FGC23742

It has been found in BigY of 7 men, interesting enough it's upstream of A1206. A1206 has shown up in number of BigY associated with men with West of Ireland heritage, when it came to running STR comparison these men often showed up close to A259+/A260+ men (though obviously A259-). Ideally what we need to do is get all the Shaughnessy's in FTDNA to test for the M222 bundle (if they are predicted M222) and than to test for likes of FGC23742 (which I just submitted to FTDNA so hopefully they offer it as a single SNP).

I should note that the main O'Shaughnessy lineage has probably been extinct since 18th century. However what's interesting about it is that it's one of few names (Concannon would be another) which only arose once in Ireland. So if over time we get a couple of dozen O'Shaughnessy's tested we might see pattern.

On a Hiberno-English note, in Ireland we always pronounce the "gh" (like a ck eg. O'Shock-nessy) in name, Shaunessy in comparison is an "americanism" when it comes to pronunciation/spelling.

This is reflective of Irish language prononunciation of Ó Seachnasaigh (s beside a "e" or "i" == sh / broad "ch" like in German "Bach" etc.)

AtWhatCost
11-23-2015, 07:30 PM
How many of this surname are we talking about, I see 1 O'Shaughnessy in the M222 project so across all projects, how many have been found/tested?

Rory Cain
11-23-2015, 08:21 PM
As I've pointed out in previous posts this one A259+/A260+ Shaughnessy is at quite a genetic distance from other M222+ O'Shaughnessy's....

I should note that the main O'Shaughnessy lineage has probably been extinct since 18th century. However what's interesting about it is that it's one of few names (Concannon would be another) which only arose once in Ireland. So if over time we get a couple of dozen O'Shaughnessy's tested we might see pattern.


Agreed about O'Shaughnessy's GD from the four others who made their result public. The 10 who remain private are anyone's guess. For the sane reason you mention, big GDs, I'm wondering if the O'Cleirigh and Cahill R-A260 samples with STRs closer to O'Shaughnessy might not be a better target.

Agreed that O'Shaughessy as a surname arose only in Ireland. Would that not also be true of O'Cleirigh? And O'Cahill.

O'Shaughneasy himself points out that his surname appears to have one point of origin, as a sept of the Ui Fiachra. He has a point there when one considers that many Irish surnames have multiple points of origin. I believe O'Cleirigh is also exclusively Ui Fiachra. Cahill may have multiple origins, but one such origin is Ui Fiachra. Same branch of Ui Fiachra as O'Shaughnessy if I recall, the Cenel Aodh.

The R-A260 O'Cleirighs and Cahills may be a better investment.

MacUalraig
11-23-2015, 08:49 PM
Well you could look on ysearch...

AtWhatCost
11-24-2015, 02:20 AM
Well you could look on ysearch...
That's not really indicative, most people don't upload their results to ysearch. Regardless, I did though and as suspected, there aren't very many at all and there are multiple haplotypes appearing for even such a small number of results. So, thank you for your suggestion, it seems to support that very few have actually tested thus far.

MacUalraig
11-24-2015, 08:18 AM
One good thing about ysearch is that it searches for similar spellings by default. Helpful for Gaelic names with myriad spellings.

Rory Cain
11-24-2015, 10:29 AM
One good thing about ysearch is that it searches for similar spellings by default. Helpful for Gaelic names with myriad spellings.

Good point. Perhaps that is why these guys missed the Cahill and O'Cleirigh/Cleary/Clarke R-A260
Types.

Dubhthach
11-24-2015, 01:37 PM
Cleary/Clarke can also cover a Bréifne surname, the name is fairly common as it basically just means "Clarke". I don't think I have access to his results to see his matches. In case of Cahill most of his matches are with men bearing surname O'Reilly, good possibility he falls into A883, where you also have one of the O'Flynn's (O'Flynn of Bréifne -- see article by McCotter in "Cavan History and Society")

Of course what would make difference there is he did BigY, if he then show shared SNP's with Shaughnessy (whose done BigY) there might be a point, however so far he's looking quite Bréifne in his surname matches.

There was a Mac Cathail family in Cavan supposedly, interesting when you look at that in Griffith you see 30 McCahill households in Cavan which might tie in with connections to O'Reilly etc.
http://www.irishtimes.com/ancestor/surname/index.cfm?fuseaction=Go.&Surname=McCahill&UserID=

Griffith seems to say there was 442 Clarke households in Cavan at that time.

Dubhthach
11-24-2015, 01:47 PM
How many of this surname are we talking about, I see 1 O'Shaughnessy in the M222 project so across all projects, how many have been found/tested?

As noted I know of two M222+ O'Shaughnessy's one of them is also A259+/A260+ (through bigY), there is wide GD between them at 67 markers (13), the second one (not tested for A259) has a match at 37 markers with another O'Shaughnessy which points at at least 3 M222+ O'Shaughnessy's.

Without testing these other men for A259 we won't know if the A259+/A260+ O'Shaughnessy is reflective of wider lineage or a NPE. With regard to the Uí Briúin here is a list of some surnames tested postive for A259+ (or A260)

Uí Briúin Aí
Connor
McManus
Concannon
McDonagh
Flynn

Uí Briúin Bréifne
Rourke
O'Reilly
Flynn (as per McCotter genealogy)
McGovern
Kernan
Ford
Fallon (FGC5939+)
Byrne/Beirne

Uí Briúin Seola
Flaherty
McHugh (potentially)

Clann Fergaile -- given a Uí Briúin lineage, many regard this as fanciful and historically dubious, ruled area of modern Galway city.
Halloran (FGC5939+)

All of the above are by and large also A260+ apart from McManus/Connor cluster which are A259+/A260-, the only other known A259+/A260- is a Connolly (Johnson), Connolly can also be a Roscommon surname.

AtWhatCost
11-24-2015, 05:30 PM
Good point. Perhaps that is why these guys missed the Cahill and O'Cleirigh/Cleary/Clarke R-A260
Types.

I believe, Cahill, also under-tested as a surname, only a handful that I can find and again, a mix of unrelated snps and then of course Clarke, there is an abundance of Clarke/Clark found in many snps under L21, in worthwhile numbers. Of course those found in the correct territory presently should carry more weight. So, I still don't see any connection other than finding onesy twosy matches, whereas those names are also found under other unrelated snps. Now if you get a Kilkelly, you might be getting there. Also, how about Ó hEidhin, any number of those found to match? You do realize you can't find one surname match for each surname to an snp and draw the conclusion the puzzle is complete? You do realize those same names are found under other snps as well? I just want to make sure there.

MacUalraig
11-24-2015, 07:44 PM
It's not as straightforward as a numbers game though as the passage of time can fill the surname up with NPE members with the original line fizzling out (as John McLaughlin argued, correctly or not, had happened with the O'Neills). So there is merit to examining minorities within a surname if there is a similar pattern amongst all the other surnames in a predicted branch. That isn't to dismiss your observation though, and its a good reminder that we need to always check back to the surname projects.

Rory Cain
11-24-2015, 08:31 PM
It's not as straightforward as a numbers game though as the passage of time can fill the surname up with NPE members with the original line fizzling out (as John McLaughlin argued, correctly or not, had happened with the O'Neills). So there is merit to examining minorities within a surname if there is a similar pattern amongst all the other surnames in a predicted branch. That isn't to dismiss your observation though, and its a good reminder that we need to always check back to the surname projects.

Quite so. And John McLaughlin copped plenty of flack from others for saying so. It remains a matter for debate, but John McLaughlin's case remains standing, whereas the arguments of some if those opposed to him have been exposed as largely wishful thinking.

Dubhthach has pointed out that several potentially Ui Fiachra surnames above also appear in other Septs. While Dubhthach didn't appear to recognise any of them had a potential Ui Fiachra connection (as a number do), I accept that the almost standard Irish problem of multiple origins applies here, and so I accept Dubhthach's advice as given. But also, until it's proven either way, I accept your advice above and await a firming up of whatever pattern develops out of this.

I have to say that your description neatly captures what we see with L1336 where the Corcumroe branch of Clanna Rory are buried amongst other unrelated surnames; and likewise L1403 where the Seven Septs of Laois are a stronger percentage but still only about half, as the South Irish DNA study of some years ago found with the Eoganacht Septs. Whatever the mechanism involved, your description is quite apt.

oneillabu
11-25-2015, 12:14 AM
whereas the arguments of some if those opposed to him have been exposed as largely wishful thinking.


Are these wishful thinking people categorised as anyone that dared to speak out against the the carefully orchestrated narratives that are in most cases pure garbage, a recent example is the the DF21 project who refused to list the surnames of the two new O'Neill entries into the L720 group, another O'Neill surname variation is currently awaiting 111 marker results that look very encouraging, it seems there that these project group administrators are dominated by tyrants with agendas who control the flow of information to push their own agendas which have nothing to do with the truth.

The entire S5488 group tick all the boxes for the Royal Dalriada line including the Niall line given in certain pedigrees as two Brothers Fergus and Muircheatach Mac Erc but this has been sidelined by the tyrants that determine what information will be seen and who will be allowed to test, I can honestly say that 95% of all administrators that I have had the misfortune to contact have been dishonest.

I could list all the examples here but what is the point, however in the end all people who have tested with FTDNA and others will be assigned to an SNP group and then the truth will ultimately come out, just like M222 Niall fairytale which is bordering on ludicrous at this stage, I have graphed the GD's of M222 O'Neill's, O'Donnells, Gallaghers etc and there is no possible way to reconcile any of these M222 people with the given pedigrees for Niall and his siblings, just like the DF23 Ui Maine which is more of the same.

Here is the Dictionary definition of Honesty

Honesty refers to a facet of moral character and connotes positive and virtuous attributes such as integrity, truthfulness, and straightforwardness, including straightforwardness of conduct, along with the absence of lying, cheating, theft, etc. Furthermore, honesty means being trustworthy, loyal, fair, and sincere.

AtWhatCost
11-25-2015, 12:27 AM
It's not as straightforward as a numbers game though as the passage of time can fill the surname up with NPE members with the original line fizzling out (as John McLaughlin argued, correctly or not, had happened with the O'Neills). So there is merit to examining minorities within a surname if there is a similar pattern amongst all the other surnames in a predicted branch. That isn't to dismiss your observation though, and its a good reminder that we need to always check back to the surname projects.
I agree actually and often wondered would the original line if it was a Chiefly line really be the most prolific or wouldn't it be a case where the large number of matches were actually followers of that Chief but not necessarily related. That can make things more complicated, particularly when there isn't a living descendant with a solid pedigree that can be tested.

Rory Cain
11-25-2015, 01:01 AM
Are these wishful thinking people categorised as anyone that dared to speak out against the the carefully orchestrated narratives that are in most cases pure garbage, a recent example is the the DF21 project who refused to list the surnames of the two new O'Neill entries into the L720 group,

I can't help you much in any dispute you may still have with John McLoughlin or anyone else. John put in hours of research work, argued his case well, and supported it with evidence. Thats the standard others need to reach to have credibility when disputing John. But I also know that new evidence can change things quickly, so I'm staying out of that one. Thanks for the invitation to be involved but I politely decline. I can't help you there. I'll await the umpire's decision on the disputes you raised.

Now, you mentioned two new O'Neills. That is something I can help you with. You may notice that were placed into the L720 group probably before the time you wrote this. Maybe you didnt notice that. I believe O'Neill is already listed as a surname. So we have covered that base too. The other one you are calling an O'Neill is a actually a surname from a different sept, as I understand it. Therefore I was unsure whether I would incur your wrath in disclosing that fact by listing that surname. But if that is your wish, consdier it done.

Dubhthach
11-25-2015, 07:12 AM
I agree actually and often wondered would the original line if it was a Chiefly line really be the most prolific or wouldn't it be a case where the large number of matches were actually followers of that Chief but not necessarily related. That can make things more complicated, particularly when there isn't a living descendant with a solid pedigree that can be tested.

If you were talking about Scotland you might have a point, in Ireland however things were quite different. The only people who bore an actual surname were those who belong to wider Fine or who had a "genealogy" that claimed they did.

Followers of a particular lord generally had their own surnames, as specific roles (marshall, bard, judge etc.) were hereditary. Likewise the "freeholder" followers of a Lord (who were clients and obligied to provide services etc. -- such as troops etc.) generally had their own genealogies.

In general actually what would often happen is as people fell out of the Derbfine (four generation agnatic inheritance group, where members are up to 4 generations removed from previous leader) they often ended up taking different surnames.

Leaving that aside Irish lords often were extremely fecund, given nature of Irish society during the Gaelic era (eg. divorce and no concept of illegitimacy).

Dubhthach
12-09-2015, 10:10 AM
With regards to terminology, here is Dr Cathy Swift at recent GGI conference in Dublin, giving a breakdown of terminology/structure:


https://www.youtube.com/watch?v=kRHCcNI8CI0

Rory Cain
12-19-2015, 06:55 AM
Are these wishful thinking people categorised as anyone that dared to speak out against the the carefully orchestrated narratives that are in most cases pure garbage, a recent example is the the DF21 project who refused to list the surnames of the two new O'Neill entries...

After due diligence it can be stated that no such request was ever made to any admin or co-admin of the R-DF21 Project. Furthermore the complainant has a long track record if such fabrications, several of which he has subsequently admitted. My staff stand exonerated. Such attacks are a poor reward for their honest efforts as volunteers.

fridurich
01-12-2017, 04:08 AM
Where do you all think M222 originated and what reasons do you give for this? I welcome any reasonable reply.

I have looked at all 17 pages of this thread. It's been a little over a year since anyone posted on it, but this seems the best place to ask that question.

Some might say "Yfull shows that M222 was formed about 4300 years ago, therefore it is unlikely it originated in Ireland or the British Isles. Plus we have a few M222 French and Germans, and they could be the remnants of the original M222 group who stayed, and they may have lived in some of the Celtic La Tene culture areas such as in southern Germany, eastern France, or Switzerland. https://www.yfull.com/tree/R-M222/ ".

Others might say "Almost everyone who is M222 today has ancestry in Ireland, Scotland, or elsewhere in the British Isles and the miniscule number of continental Europeans who have it, could be descendants of Irish mercenaries in the 17th century, and other Irish or Scottish soldiers/adventurers at about that time, or even before that time. If M222 formed in continental Europe, then why does there seem to be such a tiny amount of it there? If you are telling me that all, or almost all M222 men packed their bags and immigrated to Ireland and the British Isles, this doesn't seem very likely, really, ALL of them? If M222 originated in continental Europe, there should be way more M222 descendants there now than what we see."

So, I don't know which view is right, but look forward to your responses. Thanks in advance!

RGM
01-17-2017, 09:08 PM
I'm not sure an argument for a mainland M222 origin exists. You can ignore M222 altogether and look a couple branches higher - the "brother" or "cousin" branches of M222 are also almost entirely Irish or British.


Some might say "Yfull shows that M222 was formed about 4300 years ago, therefore it is unlikely it originated in Ireland or the British Isles.

I know you're trying to represent an argument here and this isn't meant to be your opinion, but I don't know what age would have to do with it.

fridurich
01-18-2017, 02:52 AM
I'm not sure an argument for a mainland M222 origin exists. You can ignore M222 altogether and look a couple branches higher - the "brother" or "cousin" branches of M222 are also almost entirely Irish or British.

I know you're trying to represent an argument here and this isn't meant to be your opinion, but I don't know what age would have to do with it.


I agree with you, I don't know what age would have to do with it, but I'm kind of an amateur at this stuff compared to a lot of people. I think you made a good point about the brother or cousin branches of M222 being almost entirely Irish or British. To me the thing that makes it look the most like M222 originated in the Isles or Ireland is the huge number of M222 people who have ancestors from there and the paucity of M222 people who have ancestors elsewhere.

There is someone I know of who doesn't believe that M222 existed in Ireland or the British Isles 4000 years ago. I highly respect this person and they know a lot about genetics, genetic genealogy and M222.

If you go here http://www.kennedydna.com/M222.pdf, if I understand this right, there appears to be a phylo-equivalent block (S629, S634, S635, etc.) that M222 sits among. It is puzzling to me that as far as I know, there are no descendants that are derived for this phylo-equivalent block, but negative for M222. Also, puzzling is the lone M222+ and S7073- person, J.J. Kennedy. Seems like there would be other men some where that are M222 but not derived for S7073. All other known tested M222 men are positive for S7073!

At YFull, https://www.yfull.com/tree/R-M222/ they may have some different names for some of the snps than the previous chart, and/or more phylo-equivalent snps to M222 than what is shown in the first web link I gave. At YFull, they appear to be saying M222 is about 4300 years old, yet the TMRCA is only about 1900 years ago. If I understand that right, that means that ALL M222 men will have a common ancestor only about 1900 years ago, when the age of M222 is about 4300 years old. Is it just me, or does it seem strange that for a haplogroup that is 4300 years old that the TMRCA isn't closer up towards 4300 years ago? If someone can explain that to me, in more layman's terms than extremely technical, I would be grateful.

In previous pages on this post I have seen where some people mention bottlenecks or are adamant that the tree structure of M222 does not reflect a bottleneck.

MacUalraig
01-18-2017, 07:32 AM
I'm not sure an argument for a mainland M222 origin exists. You can ignore M222 altogether and look a couple branches higher - the "brother" or "cousin" branches of M222 are also almost entirely Irish or British.



I know you're trying to represent an argument here and this isn't meant to be your opinion, but I don't know what age would have to do with it.

Age might come into it if the proposed age predates current theories on BI arrival of upstream haplogroups. I don't believe that is the case here anyway, though.

Jessie
01-18-2017, 09:32 AM
I thought M222 was only about 2,000 years old (approx) so don't think it was around 4,000 years ago. Not an expert though and I'm sure someone will give a more definitive answer.

Helgenes50
01-18-2017, 09:56 AM
I thought M222 was only about 2,000 years old (approx) so don't think it was around 4,000 years ago. Not an expert though and I'm sure someone will give a more definitive answer.

You are probably right.
For YFull, the common ancestor is 1900 years old
13617

Dubhthach
01-18-2017, 10:03 AM
I thought M222 was only about 2,000 years old (approx) so don't think it was around 4,000 years ago. Not an expert though and I'm sure someone will give a more definitive answer.

Well 1700-2000 years is the TMRCA (time to most recent common ancestor) for M222 eg. all currently tested men who are M222+ seem to share common ancestor in that period, however if you notice all M222+ men are also positive for about 30-40 other SNP's.

Now it's possible that several of these SNP's occurred together or perhaps in a short time period. But it's looking like the lineage itself branched from Z2961 anytime up to 4,000 years ago.

YFULL have following estimates:
R-Z2961 Z2961 formed 4300 ybp, TMRCA 4300 ybp
----- R-M222 Z2964 * Z2972/S640 * Z2965/S6155+35 SNPs formed 4300 ybp, TMRCA 1900 ybp

Now we have no idea when M222 might have arisen within this block of SNP's (~40) it might have mutated early, or it might have been one of the last in the block. The only way we'll ever know is if someone who is M222- is sequenced and they share some of the mutations currently seen at the M222+ level. This would lead to discovery of a very early branch lineage etc.

What seems to be the case in general is that M222 is heavily dominated by branches of DF105, well over 80%+ of men who've done M222 bundle are DF105+ -- this sub-lineage underwent rapid expansion/differentiation so much that we know of at least 15 sub-branches (S588, DF85, A259 etc.)

With regards to continental results. A swedish poster on the L513 yahoo group send a mail the other day about the Swedish "1000 genomes" project which has a web interface. You can search it for specific SNP's. I check for M222 it showed up in 6 out of 528 participants (male sample set -- Swedes appear to have spilt their project 50/50 gender wise)

All 6 were DF105+, 2 were S588+ whereas 4 fell under A1206 -- specifically FGC23739
R-A1206 A1206 formed 1250 ybp, TMRCA 1050 ybp
--- R-FGC23739 FGC23739formed 1050 ybp, TMRCA 1050 ybpinfo

http://www.ytree.net/DisplayTree.php?blockID=1119&star=false

If you ask me givin the high levels of S588 in Ireland, that and we see Morrow/Murphy (both potential angliscations of same name) under both A1206 and it's subclade FG23739. My feeling is that this Swedish M222 has origins in the Viking era. (the YFULL dates make sense)

6 out of 528 men is 1.13% of sample, as a comparison Busby had M222 showing up at level of 1.4% in sample set of 139 from Malmö

MacUalraig
01-18-2017, 12:24 PM
I thought M222 was only about 2,000 years old (approx) so don't think it was around 4,000 years ago. Not an expert though and I'm sure someone will give a more definitive answer.

Part of the problem is the way YFull added dating to their branches. The first date is just pulled down from the branch above and the second one is the TMRCA of modern descendants and is the better one to focus on.

fridurich
01-18-2017, 06:49 PM
Well 1700-2000 years is the TMRCA (time to most recent common ancestor) for M222 eg. all currently tested men who are M222+ seem to share common ancestor in that period, however if you notice all M222+ men are also positive for about 30-40 other SNP's.

Now it's possible that several of these SNP's occurred together or perhaps in a short time period. But it's looking like the lineage itself branched from Z2961 anytime up to 4,000 years ago.

YFULL have following estimates:
R-Z2961 Z2961 formed 4300 ybp, TMRCA 4300 ybp
----- R-M222 Z2964 * Z2972/S640 * Z2965/S6155+35 SNPs formed 4300 ybp, TMRCA 1900 ybp

Now we have no idea when M222 might have arisen within this block of SNP's (~40) it might have mutated early, or it might have been one of the last in the block. The only way we'll ever know is if someone who is M222- is sequenced and they share some of the mutations currently seen at the M222+ level. This would lead to discovery of a very early branch lineage etc.

What seems to be the case in general is that M222 is heavily dominated by branches of DF105, well over 80%+ of men who've done M222 bundle are DF105+ -- this sub-lineage underwent rapid expansion/differentiation so much that we know of at least 15 sub-branches (S588, DF85, A259 etc.)

With regards to continental results. A swedish poster on the L513 yahoo group send a mail the other day about the Swedish "1000 genomes" project which has a web interface. You can search it for specific SNP's. I check for M222 it showed up in 6 out of 528 participants (male sample set -- Swedes appear to have spilt their project 50/50 gender wise)

All 6 were DF105+, 2 were S588+ whereas 4 fell under A1206 -- specifically FGC23739
R-A1206 A1206 formed 1250 ybp, TMRCA 1050 ybp
--- R-FGC23739 FGC23739formed 1050 ybp, TMRCA 1050 ybpinfo

http://www.ytree.net/DisplayTree.php?blockID=1119&star=false

If you ask me givin the high levels of S588 in Ireland, that and we see Morrow/Murphy (both potential angliscations of same name) under both A1206 and it's subclade FG23739. My feeling is that this Swedish M222 has origins in the Viking era. (the YFULL dates make sense)

6 out of 528 men is 1.13% of sample, as a comparison Busby had M222 showing up at level of 1.4% in sample set of 139 from Malmö

Do you feel that the Swedish DF105 guys were probably descended from Irish men taken captive by the Vikings to Norway, and over time, their descendants moved to Sweden? Since it appears that there weren't any, or very few Swedish excursions to Ireland or the British Isles, as they seem to mainly have focused on the Baltic areas, Russia, etc.

When you mention, "Now we have no idea when M222 might have arisen within this block of SNP's (~40) it might have mutated early, or it might have been one of the last in the block. The only way we'll ever know is if someone who is M222- is sequenced and they share some of the mutations currently seen at the M222+ level. This would lead to discovery of a very early branch lineage etc.", in my opinion, it seems like it is more likely some of those SNPs at the M222 level happened together, or in a short time frame, or M222 arose towards the end of that 40 or so block of SNPs, that is closer in time to the present, or all of the above. For one thing it would make the TMRCA of about 1900 years for all known M222 descendants to make more sense. But I'm not an expert.

However, I agree probably the only way we will ever know if someone is M222-, but has some of the other 40 or so SNPs at the M222 level, is if somebody who is tested turns up negative for M222, but positive for some of the other ~40 SNPs it sits among.

fridurich
01-18-2017, 11:29 PM
Sorry, duplicate.

Jessie
01-19-2017, 09:05 AM
Do you feel that the Swedish DF105 guys were probably descended from Irish men taken captive by the Vikings to Norway, and over time, their descendants moved to Sweden? Since it appears that there weren't any, or very few Swedish excursions to Ireland or the British Isles, as they seem to mainly have focused on the Baltic areas, Russia, etc.

When you mention, "Now we have no idea when M222 might have arisen within this block of SNP's (~40) it might have mutated early, or it might have been one of the last in the block. The only way we'll ever know is if someone who is M222- is sequenced and they share some of the mutations currently seen at the M222+ level. This would lead to discovery of a very early branch lineage etc.", in my opinion, it seems like it is more likely some of those SNPs at the M222 level happened together, or in a short time frame, or M222 arose towards the end of that 40 or so block of SNPs, that is closer in time to the present, or all of the above. For one thing it would make the TMRCA of about 1900 years for all known M222 descendants to make more sense. But I'm not an expert.

However, I agree probably the only way we will ever know if someone is M222-, but has some of the other 40 or so SNPs at the M222 level, is if somebody who is tested turns up negative for M222, but positive for some of the other ~40 SNPs it sits among.

I don't necessarily think they were "captive". They could have been just Norse Gaels. This is what I think is most likely.

RGM
01-19-2017, 05:36 PM
in my opinion, it seems like it is more likely some of those SNPs at the M222 level happened together, or in a short time frame, or M222 arose towards the end of that 40 or so block of SNPs, that is closer in time to the present, or all of the above. For one thing it would make the TMRCA of about 1900 years for all known M222 descendants to make more sense. But I'm not an expert.

I'm not sure what you mean here. It makes no difference when M222 occurred because M222 is currently equivalent to all of the other SNPs in that block. Knowing where it fits in wouldn't affect the TMRCA to the block or affect the estimate of when the block originated. Unless the block is broken up, the specifics of the individual SNPs are irrelevant. It's the entire block as a whole we're interested in and we just use M222 to reference it.

Dubhthach
01-19-2017, 06:04 PM
Do you feel that the Swedish DF105 guys were probably descended from Irish men taken captive by the Vikings to Norway, and over time, their descendants moved to Sweden? Since it appears that there weren't any, or very few Swedish excursions to Ireland or the British Isles, as they seem to mainly have focused on the Baltic areas, Russia, etc.

When you mention, "Now we have no idea when M222 might have arisen within this block of SNP's (~40) it might have mutated early, or it might have been one of the last in the block. The only way we'll ever know is if someone who is M222- is sequenced and they share some of the mutations currently seen at the M222+ level. This would lead to discovery of a very early branch lineage etc.", in my opinion, it seems like it is more likely some of those SNPs at the M222 level happened together, or in a short time frame, or M222 arose towards the end of that 40 or so block of SNPs, that is closer in time to the present, or all of the above. For one thing it would make the TMRCA of about 1900 years for all known M222 descendants to make more sense. But I'm not an expert.

However, I agree probably the only way we will ever know if someone is M222-, but has some of the other 40 or so SNPs at the M222 level, is if somebody who is tested turns up negative for M222, but positive for some of the other ~40 SNPs it sits among.

Who knows, S588 after all is also found in men with Scottish surnames (other than implication it might be linked to the Cenél nEogain) -- in some stats I saw from ScotlandsDNA S588 was heavily concentrated within Ulster -- where there wasn't any Scandinavian settlements (mainly as they kept loosing battles against Northern Uí Néill -- it's quite possible that we are looking at movement connected to the GallGhaeil (Norse-Gaels). Leaving that aside there has course been internal migration within Scandinavia over the last 1000 years.

What I find interesting is that the Swedish A1206 all appears to fall into it's own sub-branch.

As for the M222 block of SNP's. Well sure several of them could have occurred together, I imagine if the lineage spilt from Z2961 4000 years ago and than only started expanding around 1700 years ago. That gives a 2300 year period where these SNP's could have arisen.

Dubhthach
01-19-2017, 06:07 PM
What I would say about the actual block of SNP's. Is that if we could find a lineage that breaks it (such a theoretical lineage might not fit the STR cluster) this provides us with an extra datapoint. So for example if such a theoretical lineage spilt 2,500 years ago and was found in men with purely a northern England origin. This could imply that the history of lineage was tied to Northern Britain and from there expanded into Ireland (post 2,500 years ago). In which case you could theorise that such expansion was to do with renewed contacts between Northern half of Ireland and Northern Britain in period after 300BC (as seen in archaeological record)

jdean
01-19-2017, 06:48 PM
With the age of the M222 block it's quite possible bits of it'll turn up in bronze age aDNA samples, fingers crossed : )

dp
01-19-2017, 10:23 PM
Get them shovels and spades out...
dp

fridurich
01-20-2017, 04:11 AM
I don't necessarily think they were "captive". They could have been just Norse Gaels. This is what I think is most likely.

Yes, you are right they could have been Norse Gaels. I didn't think about that.

fridurich
01-20-2017, 04:36 AM
What I would say about the actual block of SNP's. Is that if we could find a lineage that breaks it (such a theoretical lineage might not fit the STR cluster) this provides us with an extra datapoint. So for example if such a theoretical lineage spilt 2,500 years ago and was found in men with purely a northern England origin. This could imply that the history of lineage was tied to Northern Britain and from there expanded into Ireland (post 2,500 years ago). In which case you could theorise that such expansion was to do with renewed contacts between Northern half of Ireland and Northern Britain in period after 300BC (as seen in archaeological record)

It would be so nice if we found a substantial number of ancient skeletons that had the M222 block in Ireland or the British Isles, or, if like you mentioned earlier, we found a lineage that could break that block.